Cross-Cultural Pragmatics

W
DE
G
Cross-Cultural Pragmatics
The Semantics of Human Interaction
Second edition
by
Anna Wierzbicka
Mouton de Gruyter
Berlin New York 2003
Mouton de Gruyter (formerly Mouton, The Hague)
is a Division of Walter de Gruyter GmbH & Co. KG, Berlin.
The first edition was published in 1991 as volume 53
of the series Trends in Linguistics. Studies and Monographs
@ Printed on acid-free paper which falls
within the guidelines of the ANSI to ensure
permanence and durability.
ISBN 3-11-017769-2
Bibliographic information published by Die Deutsche Bibliothek
Die Deutsche Bibliothek lists this publication in the Deutsche
Nationalbibliografie; detailed bibliographic data is available in the
Internet at <http://dnb.ddb.de>.
Copyright 1991, 2003 by Walter de Gruyter GmbH & Co. KG, 10785 Berlin
All rights reserved, including those of translation into foreign languages. No part of this
book may be reproduced or transmitted in any form or by any means, electronic or mechan-
ical, including photocopy, recording, or any information storage and retrieval system, with-
out permission in writing from the publisher.
Printing: Gerike GmbH, Berlin. - Binding: Liideritz & Bauer-GmbH, Berlin.
Cover design: Sigurd Wendland, Berlin.
Printed in Germany.
Introduction to the second edition
I am very happy to see the demand for a new edition of my 1991 Cross-
Cultural Pragmatics - the Semantics of Human Interaction. I am also
happy to be able to say, in 2003, that since this book was first published
the field of cross-cultural pragmatics has advanced enormously; and
furthermore, that this progress has not only not made my 1991 Cross-
Cultural Pragmatics dated, but that, on the contrary, its tenets and its
overall approach have been essentially vindicated. A decade ago, the
"pragmatic" scene was still largely dominated by the search for the "uni-
versals of politeness" and for the "universal maxims of conversation".
The widely accepted paradigms were those of Brown and Levinson's
(1978, 1987) theory of politeness, which affirmed "pan-cultural interpret-
ability of politeness phenomena" (1978: 288), and Grice's (1975) theory
of conversation, which posited a number of universal conversational
principles. It is heartening to see to what extent the situation has now
changed.
In the nineteen eighties, and well into the nineties, the idea that inter-
personal interaction is governed, to a large extent, by norms which are
culture-specific and which reflect cultural values cherished by a particular
society went against the grain of what was generally accepted at the time,
and successive conferences of the Inernational Pragmatic Association and
other similar occasions were dominated by studies seeking to confirm
Grice's "maxims" and Brown and Levinson's "universals of politeness"
in this or that new area, and this or that new language.
In 1978, Brown and Levinson set out "to describe and account for
what is in the light of current theory a most remarkable phenomenon.
This is the extraordinary parallelism in the linguistic minutiae of the ut-
terances with which people choose to express themselves in quite unre-
lated languages and cultures" (Brown and Levinson 1978: 60). A quarter
of a century later, it is increasingly widely accepted that this "extraordi-
nary parallelism" was largely an illusion due to that "light of current
theory". (If you set out to show that everything can be described in terms
of "negative and positive face" you may indeed find that everything can
be so described.) What is seen as more remarkable today, is the extent of
cross-linguistic and cross-cultural differences in ways of speaking. Brown
and Levinson (1978: 61) described it as their goal "to rebut the once-fash-
VI Introduction to the second edition
ionable doctrine of cultural relativity in the field of interaction" and "to
show that superficial diversities can emerge from underlying universal
principles and are satisfactorily accounted for only in relation to them".
Their major conclusion was that "interactional systematics are based
largely on universal principles" (1978: 288). Today, it is increasingly ac-
cepted that those diversities in ways of speaking and interacting are not
superficial at all and that they can be accounted for, above all, in terms
of different cultural attitudes and values; and the "cultural relativity in
the field of interaction" is increasingly seen as a reality and an important
subject for investigation.
When in 1983 I presented, at the monthly meeting of the Sydney Lin-
guistic Circle, a paper entitled "Different cultures, different languages,
different speech acts: English vs. Polish" (Wierzbicka 1985), in which I
argued that the supposedly universal maxims and principles of "polite-
ness" were in fact rooted in Anglo culture, my ideas were regarded as
heretical. When I argued, in particular, that the "freedom from imposi-
tion", which Brown and Levinson (1978: 66) saw as one of the most
important guiding principles of human interaction, was in fact an Anglo
cultural value, and that the avoidance of "flat imperative sentences",
which Searle (1975: 69) attributed to the "ordinary (human, A. W.) con-
versational requirements of politeness", did not reflect "universal prin-
ciples of politeness" but rather, expressed special concerns of modern
Anglo culture, my claims were confidently dismissed. As a matter of fact,
it was the hostile and dismissive reaction of that audience which was for
me the initial stimulus for engaging in a long-term campaign against what
I saw as a misguided orthodoxy of that time.
From the perspective of the intervening years, I must be grateful for
the negative reaction of that Sydney audience to a paper which became
the nucleus of my 1991 Cross-Cultural Pragmatics. I am even more grate-
ful, however, to other linguists, who in that inhospitable post-Gricean
climate were also raising their voices in defence of culture as a key factor
determining ways of speaking, and in particular, to those who ventured
to link language-specific ways of speaking with different cultural values.
To mention just a few scholars, whom I saw in those early years, and
whom I see now, as "comrades-in-arms": Ho-min Sohn, the author of
a pioneering study "Intercultural communication and cognitive values"
(1983); Tamar Katriel, the author of Talking Straight: Dugri speech in
Israeli Sabra culture (1986); Yoshiko Matsumoto, the author of "Reexam-
ination of the universality of face: politeness phenomena in Japanese"
(1988); James Matisoff, the author of Blessings, Curses, Hopes, and Fears:
Introduction to the second edition VB
Psycho-Ostensive Expressions in Yiddish (1979); Thomas Kochman, the
author of Black and White Styles in Conflict (1981); Sachiko Ide, the
author of a study on the Japanese value of wakimae or discernment
(1989); Donal Carbaugh, the author of Talking American (1990); and
closer to home, my colleagues: Cliff Goddard, whose numerous publica-
tions are listed in the References; Jean Harkins, the author of Bridging
Two Worlds: Aboriginal English and Cross-Cultural Understanding (1994);
Felix Ameka, the author of studies on Ghanaian conversational routines
and the editor of a volume on interjections (see the References); and
Michael Clyne, the author of Intercultural Communication at Work: Cul-
tural Values in Discourse (1994). Last but not least, I would like to men-
tion the important role of two open-minded and cross-culturally alive
journals: Jacob Mey's Journal of Pragmatics, and Marcelo Dascal's Prag-
matics and Cognition.
Outside linguistics, there were of course anthropologists who did not
give in to the superficial and anti-cultural universalism of the time and
who continued to focus on the language-particulars and to probe the
links between ways of speaking, ways of thinking, ways of feeling and
ways of living. To mention just a few names and works, particularly
important from a linguistic point of view: Catherine Lutz, the author of
the classic book Unnatural Emotions: Everyday sentiments on a Microne-
sian atoll and their challenge to Western theory (1988); Richard Shweder,
the founder of "cultural psychology" and the author of Thinking Through
Cultures - Expeditions in Cultural Psychology (1991); Dorothy Holland
and Naomi Quinn, the editors of Cultural Models in Language and
Thought (1987); and Roy D'Andrade and Claudia Strauss, the authors of
Human Motives and Cultural Models (1992). I would also like to mention
here two journals which I see as especially important: Ethos and Culture
and Psychology.
There were also some philosophers who started to question the prag-
matic theories of Grice, Griceans, and "neo-Griceans" from a philosophi-
cal as well as cross-linguistic point of view. In particular, Wayne Davis
(1998) has argued in a book-length critique that "the Gricean theory has
been barren" and that "the illusion of understanding provided by the
Gricean theory has only served to stifle inquiry" (Davis 1998: 3). "The
Gricean explanation of common implicatures" is, Davis argued, "un-
dermined by the existence of nonuniversal implicature conventions"
(Davis 1998: 183).
For example, Grice and his followers (e.g. Levinson 1983) have
claimed that the correct interpretation of a tautology like War is war can
VIll Introduction to the second edition
be calculated from some universal maxims of conversation. Davis points
out (with reference to Chapter 10 of my Cross-Cultural Pragmatics) that
this claim is refuted by the observation that such tautologies receive dif-
ferent interpretations in different cultures, and he concludes: "The moral
is clear. Generalized tautology implicatures ... are not explained by Gri-
cean Maxims" (Davis 1998: 46). In a similar context, Davis (1998: 168)
quotes and endorses my own observation that "from the outset, studies
in speech acts have suffered from an astonishing ethnocentrism" (Wierz-
bicka 1985: 145).
Since the decline of the Gricean paradigm, which, as Davis puts it, has
only served to stifle inquiry, defines to a large extent the difference in the
context between this second edition of my Cross-Cultural Pragmatics and
the 1991 one, I hope I can be forgiven for quoting at some length Davis'
historical account, including his comments on my own work.
Many of the criticisms I present have been known for some time. But the
import and seriousness of the defects individually and collectively have not
been widely appreciated, and the problems have had little impact on the gene-
ral acceptance of Gricean theory. The best known critics of the Gricean theory
have either expressed confidence that solutions would be found within the
Gricean framework (Harnish 1976) or presented alternative theories with sim-
ilar defects (Sperber and Wilson 1986).... Only one author (Wierzbicka 1987)
has argued that the conception is fundamentally flawed. (Davis 1998: 3)
I hasten to add that Davis has reserved some critical comments for me
too, and that I will quote these later. What matters at this point is the
historical record, which the reader of this second edition is entitled to
know.
From the historical, as well as theoretical, point of view, it is important
to note that a powerful impulse for the rise of cross-cultural pragmatics
in the last decade came from the growing field of studies focussed on
cross-cultural (or inter-cultural) communication. I have quoted in the
1991 book Deborah Tannen's (1986: 30) statement that "the future of the
earth depends on cross-cultural communication". At a time when every
year millions of people cross the borders, not only between countries but
also between languages, and when more and more people of many dif-
ferent cultural backgrounds have to live together in modern multi-ethnic
and multi-cultural societies., it is increasingly evident that research into
differences between cultural norms associated with different languages is
essential for peaceful co-existence, mutual tolerance, necessary under-
standing in the work-place and in other walks of life in the increasingly
"global" and yet in many places increasingly diversified world.
Introduction to the second edition IX
The once popular assumption that the "principles of politeness" are
essentially the same everywhere and can be described in terms of "univer-
sal maxims" such as those listed in Leech (1983: 132) flies in the face of
reality as experienced by millions of ordinary people - refugees, immi-
grants, the children of immigrants, caught between their parents and the
society at large, cross-cultural families and their children, and also by
monolingual "stay-at-homes" who suddenly find themselves living in so-
cieties which are ethnically, culturally and linguistically diverse.
In addition to their obvious untruth in relation to daily experiences of
millions of people, the supposed "universals of politeness" and the sup-
posed "universal principles of conversation" are clearly of no use in the
practical task of furthering cross-cultural communication. When, for ex-
ample, a well-meaning, liberal Anglo-Australian says of her Chinese
neighbours that "they are very good neighbours - but they are so rude ...
for example, they said to me: cut down that branch - we don't want it
on our side of the fence" (Canberra 2002), if we as linguists tell her and
others like her that the principles of politeness are essentially the same
everywhere (recall Brown and Levinson's tenet of "pan-cultural interpret-
ability of politeness phenomena" quoted earlier), we can only confirm her
in her view that the Chinese neighbours are very rude (cf. Clyne 1994).
The tremendous practical importance of identifying, and describing,
the culture-specific norms of "politeness" and, more generally, norms of
interpersonal interaction, has been increasingly recognized by the field of
language teaching. In this field, too, the realization grew steadily over
the last decade or so that "Grice's Razor", which extols the economical
virtues of concentrating on the supposed universality of the "underlying
principles" and which cuts off "unnecessary" culture-specific explana-
tions, spells out a disaster for the students' communicative competence
and their ability to survive socially in the milieu of their "other" language.
As Kramsch (1993) puts it in her book Context and Culture in Language
Teaching:
If ... language is seen as social practice, culture becomes the very core of
language teaching. Cultural awareness must then be viewed both as enabling
language proficiency and as being the outcome of reflection on language profi-
ciency.... Once we recognize that language use is indissociable from the cre-
ation and transmission of culture, we have to deal with a variety of cultures.
(Kramsch 1993: 89)
A key question for Kramsch and many other contemporary theorists and
practitioners of language teaching aimed at communicative competence
is this: "How can a foreign way of viewing the world be taught via an
x Introduction to the second edition
educational culture which is itself the product of native conceptions and
values?" (Kramsch 1993: 9). Clearly, it is not a question that Grice's
Razor or the supposedly universal notion of "positive" or "negative"
"face" can help answer.
My own long campaign against the fictitious and harmful "universals
of politeness" and "universal principles of human conversation" is rooted
in my own experience as a "language migrant" (to use a term introduced
by Mary Besemeres, 1998 and 2002) - from Polish into English, especially
academic English, in which I have written many books and articles, and
also, into Australian English, which has been my daily linguistic environ-
ment for thirty years. I have described this experience in some detail in
an article entitled "The double life of a bilingual - a cross-cultural per-
spective" (Wierzbicka 1997b). On a very small scale, this article illustrates
an important new aspect of cross-cultural pragmatics as it has evolved
over the last decade or so: the new alliance between, on the one hand,
linguistic pragmatics, based on "hard linguistic evidence" and rigorous
linguistic analysis, and, on the other, the new field of study focused on
the "soft data" of personal experience of cross-cultural and cross-linguis-
tic living (cf. Besemeres 2002; Dalziell 2002).
I have referred to my own cross-linguistic and cross-cultural experience
in a number of publications, both before and after the 1991 Cross-Cul-
tural Pragmatics. Here, I will permit myself to adduce several long quotes
from that 1997 cross-cultural memoir, which deliberately takes a personal
rather than "objective" perspective. I believe that such a personal per-
spective legitimizes the insistence with which proponents of cross-cultural
pragmatics have been challenging, in the last decade or so, the earlier
paradigm. Commenting on my life in Australia, to which I emigrated
from Poland in 1972 (having married an Australian) I wrote:
I had to start learning new "cultural scripts" to live by, and in the process I
became aware of the old "cultural scripts" which had governed my life hith-
erto. I also became aware, in the process, of the reality of "cultural scripts"
and their importance to the way one lives one's life, to the image one projects,
and even to one's personal identity.
For example, when I was talking on the phone, from Australia, to my
mother in Poland (15,000 km away), with my voice loud and excited, carrying
much further than is customary in an Anglo conversation, my husband would
signal to me: 'Don't shout!' For a long time, this perplexed and confused me:
to me, this 'shouting' and this 'excitement' was an inherent part of my person-
ality. Gradually, I came to realise that this very personality was in part cultur-
ally constituted. (Wierzbicka 1997b: 119)
Introduction to the second edition Xl
The realization of the close links betweeen my ways of speaking, my
personality and my Polishness raised for me the question that countless
other immigrants are constantly confronted with: to what extent was it
desirable, or necessary, to change myself in deference to my new cul-
tural context?
Early in our life together, my husband objected to my too frequent - in his
view - use of the expression of course. At first, this puzzled me, but eventually
it dawned on me that using of course as broadly as its Polish counterpart
oczywiscie is normally used would imply that the interlocutor has overlooked
something obvious. In the Polish 'confrontational' style of interaction such an
implication is perfectly acceptable, and it is fully consistent with the use of
such conversational particles such as, for example, przeciei ('but obviously -
can't you see?'). In mainstream Anglo culture, however, there is much more
emphasis on 'tact', on avoiding direct clashes, and there are hardly any con-
frontational particles comprarable with those mentioned above. Of course does
exist, but even oj' course tends to be used more in agreement than in dis-
agreement (e. g. 'Could you do X for me?' - 'Of course'). Years later, my
bilingual daughter Mary told me that the Polish conversational expression alei
oczytviscie: 'but-EMPHATIC of course' (which I would often replicate in English
as 'but of course') struck her as especially 'foreign' from an Anglo cultural
point of view; and my close friend and collaborator Cliff Goddard pointed
out, tongue in cheek, that my most common way of addressing him (in Eng-
lish) was 'But Cliff ... '. (Wierzbicka 1997b: 119)
Thus, I had to learn to avoid overusing not only of course but also many
other expressions dictated by my Polish "cultural scripts"; and in my
working life at an Anglo university this restraint proved invaluable, in-
deed essential.
I had to learn to 'calm down', to become less 'sharp' and less 'blunt', less
'excitable', less 'extreme' in my judgements, more 'tactful' in their expression.
I had to learn the use of Anglo understatement (instead of more hyperbolic
and more emphatic Polish ways of speaking). I had to avoid sounding 'dog-
matic', 'argumentative', 'emotional'. (There were lapses, of course.) Like the
Polish-American writer Eva Hoffman (1989) I had to learn the use of English
expressions such as 'on the one hand ... , on the other hand', 'well yes', 'well
no', or 'that's true, but on the other hand'.
Thus, I was learning new ways of speaking, new patterns of communica-
tion, new modes of social interaction. I was learning the Anglo rules of turn-
taking ('let me finish!', 'I haven't finished yet!'). I was learning not to use the
imperative ('Do X!') in my daily interaction with people and to replace it with
a broad range of interrogative devices ('Would you do X?' 'Could you do X?'
'Would you mind doing X?' 'How about doing X?' 'Why don't you do X?'
'Why not do X?', and so on). (Wierzbicka 1997b: 119-120)
xu Introduction to the second edition
As I discussed in that 1997 memoir, these were not just changes in the
patterns of communication, these were also change 'in my personality. I
was becoming a different person, both in the context of my cross-cultural
family and in the context of my work as a university teacher.
Students' course assessment questionnaires have often thrown light on my
cultural dilemmas. Thus, while often very positive and praising my 'enthusi-
asm', for a long time they also often included critical accents referring to my
'intensity', 'passion" and 'lack of detachment'. I was coming from a language
and culture system (Polish) where the very word beznamif2tny (lit. 'dispassion-
ate') has negative connotations, but I was lecturing in a language (English)
where the word dispassionate implies praise while the word emotional has nega-
tive connotations. I had to learn, then, to lecture more like a 'spokesman' and
less like an 'advocate' (in Kochman 1981 terms). I had to learn to become less
'emotional' and more 'dispassionate' (at least in public speaking and in aca-
demic writing). (Wierzbicka 1997b: 120)
And yet, while I saw some cultural adaptation as necessary I did not
want to adapt too much; I felt instinctively that the social benefits of such
an adaptation needed to be balanced against the personal cost involved in
it.
There were therefore limits to my malleability as a 'culturally consti-
tuted self'. There were English modes of interaction that I never learnt
to use - because I couldn't and because I wouldn't: they went too much
against the grain of that 'culturally constituted self'. For example, there
was the 'How are you' game: 'How are you?' - 'I'm fine, how are you?';
there were weather-related conversational openings ('Lovely day isn't
it?' - 'Isn't it beautiful?'). There were also 'white lies' and 'small talk'
(the latter celebrated in a poem by the Polish poet and professor of Slavic
literatures at Harvard University, Stanislaw Baranczak).
The acute discomfort that such conversational routines were causing
me led me to understand the value attached by Polish culture to 'sponta-
neity', to saying what one really thinks, to talking about what one is
really interested in, to showing what one really feels. It also led me to
contemplate the function of such linguistic lubricants in Anglo social
interaction. Why was it that Polish had no words or expressions corre-
sponding to 'white lies' or 'small talk'? Why was it that English had no
words or expressions corresponding to basic Polish particles and 'conver-
sational signposts' such as przeciei, alei ('but can't you see?') alei skg,die
(lit. 'but where from?' i.e. where did you get that idea?), or skg,die znowu
('but where from again?') - all expressions indicating vigorous dis-
agreement, but quite acceptable in friendly interaction in Polish?
Introduction to the second edition XlII
As I meditated on my experience, and as I discussed it with other
immigrants, I developed a strong theoretical interest in the problems of
cross-cultural understanding and a -deep conviction that the universalist
theories of human interaction dominant of the time were fundamentally
flawed.
Clearly, the rules for 'friendly' and socially acceptable interaction in Polish and
in English were different. Consequently, I could never believe in the "universal
maxims of politeness" and in the universal "logic of conversation" promul-
gated in influential works such as Grice (1975), Leech (1983) or Brown and
Levinson (1978, 1987). I knew from personal experience, and from two de-
cades of meditating on that experience, that Polish "maxims of politeness"
and the Polish rules of "conversational logic" were different from the Anglo
ones. (Wierzbicka 1997b: 120)
As these quotes make clear, the personal knowledge derived from such
personal experience was not purely theoretical: above all, it was practical.
I had no doubt that the insistence on cultural differences was not only
theoretically justified (because these differences were real) but also that
acknowledging them, and above all, describing them, was vitally impor-
tant for the practical purposes of cross-cultural communication and un-
derstanding; and in the case of people like myself, of daily living.
The 1991 edition of my Cross-Cultural Pragmatics was an attempt to
challenge the Gricean and Brown-and-Levinsonian paradigms, and to
expose the anglocentric character of various supposedly universal max-
ims, principles and concepts (including the key concept of "face", which
was the linchpin of Brown and Levinson's theory of "politeness"). Twelve
years later it can be said that tide has changed and it may seem unneces-
sary and unkind to press the same charges again.
In response I would say that, first of all, many linguists who are out
of touch with the developments in the fields of intercultural communica-
tion and language teaching are not aware of this change of tide and
assume that the Gricean and neo-Gricean paradigms are still held by
"those in the know" in the same esteem as they once were. But there is
also another reason why some of the old charges still need to be pressed.
This second reason has to do with the fact that, paradoxically, while
the universalist pragmatic frameworks developed in the seventies were
gradually losing their appeal, the program of actually describing the dif-
ferent ways of speaking and thinking linked with different cultures con-
tinued to encounter a great deal of resistance and criticism.
As the differences between cultures and subcultures were increasingly
celebrated, there was also a growing suspicion of any generalizations as
XIV Introduction to the second edition
to what exactly these differences might be. Diversity was seen as beautiful
but also as inherently elusive and indescribable. With the growing empha-
sis on diversity, the view gradually developed that diversity was every-
where, and that while those differences could and should be celebrated
they could not be described. Thus, in many quarters, there developed a
great fear of the notion of culture (especially, "a culture"), and attempts
to identify any differences between particular cultures came to be seen as
"static culturologies" (cf. Darnell 1994).
For example, the anthropologist Eric Wolf, writing of "the hetero-
geneity and the historically changing interconnectedness of cultures"
(Wolf 1994: 5), argued that "notions of a common cultural structure un-
derlying all this differentiation sound a bit too much like a little cultural
homonuculeus built into everyone through the process of socialization"
(Wolf 1994: 6). Another anthropologist, Immanuel Wallenstein, spoke in
the same vein in his commentary on Wolf's paper, for example: "races,
cultures, and peoples are not essences. They have no fixed contours. They
have no self-evident content. Thus, we are all members of multiple, in-
deed myriad, 'groups' - crosscutting, overlapping, and ever-evolving"
(Wallenstein 1994: 5; for discussion, see Wierzbicka 1997a).
There can be no quarrel with the claims that "cultures are not es-
sences", that "cultures are not monads", and that "cultures have no fixed
contours". But to conclude from this that cultures cannot be discussed,
described, and compared at all - because they have no substance at all -
would be a spectacular case of throwing the baby out with the bath water.
It would also be a conclusion denying the subjective experience of immi-
grants, and, as I have argued in detail elsewhere (Wierzbicka, forthcom-
ing), one going against their vital interests. To deny the validity of the
notion of culture-specific cultural patterns (including "Anglo" cultural
patterns) is to place the values of political correctness above the interests
of socially disadvantaged individuals and groups.
At this point, it will be apposite to return to Davis' (1998) critical
comment on my own work, to which I have alluded earlier. Characteristi-
cally, this comment refers especially to my remarks on Anglo culture.
To quote:
To the extent that norms for polite, cooperative, efficient communication vary
from culture to culture, so should implicature conventions. Thus Wierzbicka
(1985) offered the "heavy restrictions on the use of the imperative in English
and the wide range of use of interrogative forms in performing acts other
than questions" as "striking linguistic reflexes" of the Anglo-Saxon cultural
tradition, one that "places special emphasis on the rights and on the autonomy
Introduction to the second edition xv
of every individual, which abhors interference in other people's affairs," and
so on. She observed that languages such as Polish, used by speakers with
opposed cultural traditions, have different conventions involving imperatives
and interrogatives. The fact that Wierzbicka is fighting ethnocentrisn1 with
cultural stereotypes does not diminish her point. (Davis 1998: 174-175)
The fear of "cultural stereotypes" has been as great an obstacle in the
development of cross-cultural pragmatics as has the fear of "essentialism"
and the "reification" of cultures. Giving in to this fear, Davis seems to
be doing something analogous to what he himself criticized Brown and
Levinson for, when he said that they "note the evidence but insist the
'underlying principles' are universal, derivable from universal face as-
sumptions and rationality" (Davis 1998: 167). Similarly, Davis notes the
evidence concerning the language-specific character of pragmatic conven-
tions but he rejects off-hand any possible links between different prag-
matic conventions and different cultural attitudes and values. He accepts
that those conventions are not universal and he himself calls for "histori-
cal and sociolinguistic research ... which did not and could not arise
when the Gricean theory held sway" (Davis 1998: 3). At the same time,
however, he feels compelled to dismiss cross-cultural generalizations as
"stereotyping".
Yet from the point of view of effective cross-cultural understanding
and intercultural comlllunication it is essential not only to know what
the conventions of a given society are but also how they are related to
cultural values. For example, the Chinese immigrants in Canberra need
to be told not only to be careful with the imperative when speaking to
their Anglo neighbours, but also, why the imperative (e. g. "cut down
that branch - we don't want it on our side of the fence") can be perceived
as offensive and "rude" in Australia. Similarly, the Anglo-Australians
need to be told not only that they should be "tolerant" to their Chinese
neighbours, but also, that their own imperative-avoiding conventions re-
flect special historically-shaped concerns of their own culture rather than
any natural and universal principles of politeness.
With the increasing domination of English in the world, both Anglos
and non-Anglos need to learn about various Anglo "cultural scripts". To
try to describe these scripts, and to explain the values reflected in them,
is not to indulge in stereotyping, but on the contrary, it is to help Anglos
to overcome their inclination to stereotype Chinese (or, for that matter,
Polish) immigrants as "rude", while at the same time helping the immi-
grants to better fit in, socially, and to improve their lives. As more and
more often noted by bilingual and bicultural theorists such as, for exam-
XVI Introduction to the second edition
pIe, Young Yun Kim (2001), for millions of people in the modern world
cultural adaptation is necessary for survival; and liberal monocultural
Anglos fixated on fighting "stereotypes" are not helping the cause of that
adaptation and of increased inter-cultural understanding.
In this context, I would like to emphasize again the new light thrown
on problems of cross-cultural pragmatics by the new field of studies
focussing on the experience of bilingual and bicultural persons, and in
particular, on the immigrant experience. It is becoming more and more
obvious to those concerned with cross-cultural understanding that in ad-
dition to objective methods usually employed in social sciences (data col-
lection, statistical tables, diagrams, and so on), the voices of flesh-and-
blood people crossing linguistic and cultural boundaries need also to be
taken into account. "The immigrant experience of having to 'translate
oneself' from one's mother tongue into a foreign language and losing
part of oneself in the process" (Besemeres 2002: 9) can expose what Davis
(1998) calls the stifling effect of universalistic accounts of human conver-
sation better than many scrupulous objective studies of linguistic compe-
tence or behaviour. It can also show more clearly than purely theoretical
debates that cultures are real and that they can influence and even shape
people's lives and people's selves. If this or that theoretical framework is
not helpful in describing cultural differences in ways of speaking, think-
ing and feeling, it can only blame itself for its irrelevance to cross-cultural
understanding, intercultural communication, language teaching, and
what John Locke called "human understanding" in general.
In my 1991 Cross-Cultural Pragmatics I did seek to describe and com-
pare different cultures, and I did use expressions like "Vietnamese cul-
ture", "Japanese culture", "Anglo-American culture", "Polish culture",
and so on. Given the potential for misunderstanding that such terms
carry with them I would now prefer to avoid them, as far as possible,
and to use instead terms like "cultural patterns", and especially, "cultural
scripts". Both these terms were used in the 1991 text of the book, but
along with, say, "Japanese cultural scripts" or "Anglo cultural scripts" I
was also using quite freely terms like "Japanese culture" or "Anglo cul-
ture". Given present-day sensitivities, it will be in order to warn the
reader explicitly that by using such terms I did not mean to imply that
I see those cultures as immutable essences, self-contained monads, or
"bounded, coherent and timeless systems of meaning" (Strauss and
Quinn 1997: 3). Rather, I was using such terms as convenient abbrevi-
ations, referring to complexes of shared understandings or, as colleagues
and I have been calling them for years, "cultural scripts". To quote
Strauss and Quinn (1997) again:
Introduction to the second edition XVll
Our experiences in our own and other societies keep reminding us that some
understandings are widely shared among members of a social group, surpris-
ingly resistant to change in the thinking of individuals, broadly applicable
across different contexts of their lives, powerfully motivating sources of their
action, and remarkably stable over succeeding generations. (Strauss and Quinn
1997: 3)
In the twelve years which have elapsed between the first and the present
edition of his book, colleagues and I have been increasingly moving from
the language of "cultures" to that of "cultural scripts". Since we have
never thought of cultures as "timeless monads", this is above all a change
in the style of exposition. The formulae included in this book under head-
ings like "Polish culture" or "Japanese culture" would now be presented
explicitly as "cultural scripts". Although this would be only a change in
presentation, not in substance, it would be an important change. Since
for logistic reasons this change is not being made in the text of this book,
the reader of this second edition is asked to bear this point in mind: this
book is not seeking to describe whole cultures, let alone to imply that
these cultures are immutable, but rather, to articulate certain specific
"cultural scripts".
At the same time, I would like to point out to the reader that since the
publication of the first edition, the idea of "cultural scripts", implicit in
this book, has come into its own as a full-fledged theory - the theory of
cultural scripts, which has by now resulted in many descriptive studies,
across many languages and cultures (or "lingua-cultures", cf. Attinasi
and Friedrich 1995). Since the idea of cultural scripts has now been devel-
oped into a theory of cross-cultural pragmatics, inter-cultural communi-
cation and indeed cross-cultural understanding in general, the reader of
this second edition may wish to follow up the development of this theory
and its applications in descriptive studies. For this, the starred references
listed at the end of this introduction may be particularly useful.
The theory of cultural scripts is an offshoot of the NSM (Natural
Semantic Metalanguage) theory, on which all the analyses in this book
are based. In a nutshell, this theory postulates that semantic analysis
should be based on empirically established universal human concepts,
that is, simple concepts realized in all languages as words or word-like
elements, such as GOOD and BAD, KNOW, THINK, WANT and SAY, DO and
HAPPEN and fifty or so others. In relation to cross-cultural pragmatics,
this means that cultural norms of speaking should be formulated neither
in technical or semi-technical English terms like "formal" and "informal"
or "direct" and "indirect", nor in terms of English folk categories like
XVl11 Introduction to the second edition
"apology", "compliment", "sarcasm", "understatement" and so on, but
rather in terms of simple words which have equivalents in all languages,
such as those mentioned above (in small capitals).
The use of such concepts can free us from what Goddard (2002c, in
press a, b, c) calls "terminological ethnocentrism" and give us a neutral,
culture-independent metalanguage for describing different cultural
norms. At the same time, the use of such concepts allows us to capture
the native speaker's point of view, without distorting it through the appli-
cation of descriptive tools rooted in the English language or Anglo aca-
demic culture.
On this point, NSM-based approach to cross-cultural pragmatics dif-
fers radically from that characteristic of works like Blum-Kulka et al.'s
(1989) Cross-Cultural Pragmatics: Requests and Apologies or Kasper and
Blum-Kulka's (1993) Interlanguage Pragmatics. While the works in this
tradition must be appreciated for their attention to cultural differences
reflected in ways of speaking, they cannot escape the charge of termino-
logical, and not only terminological, ethnocentrism. Given that words
like requests and apologies stand for conceptual artefacts of the English
language, using them as analytical tools inevitably involves imposing an
Anglo perspective on other languages and cultures. To describe ways of
speaking across languages and cultures in terms of folk categories en-
coded in English is like describing English talk in terms of Japanese,
Hebrew or Russian folk categories (e.g. the Japanese wakimae, cf. Ide
1989; the Hebrew dugri, cf. Katriel 1986; or the Russian vran'e, cf. Wierz-
bicka in press). But of course nobody would dream of describing English
in such terms.
The unshakable conviction shared by so many semanticists and prag-
maticists that it is all right to try to describe all languages through Eng-
lish terms untranslatable into the language of speakers whose ways of
thinking those terms are supposed to explain and illuminate shows the
same astonishing anglocentrism as the Gricean and post-Gricean max-
ims, principles, and "conversational postulates" (cf. Gordon and Lakoff
1975) once did. By contrast, words like good and bad or say, think, know
and want, which as evidence suggests have morpho-lexical exponents in
all languages, free us from an Anglo perspective, while allowing us at the
same time to retain a mini-lexicon of sixty or so English words as a
practical lingua franca for articulating different culture-specific conven-
tions, norms and values.
Judging from some reviews, and some other responses to the first edi-
tion of this book which have been reported to me, I was understood by
Introduction to the second edition XIX
some readers of that edition as claiming that "semantics should swallow
pragmatics". This is a misunderstanding that the theory of cultural scripts
should effectively dispel. What I did and do claim is, first, that a great
many subjective and attitudinal meanings are indeed semantically en-
coded, and second, that since all observations on language use have to
be themselves formulated in some language, their descriptive and explan-
atory power depends on the adequacy of that (meta-)language. For exam-
ple, claims that in many societies people are guided in their ways of
speaking by principles like "don't impose" or "be relevant" depend on the
English words ilnpose and relevant, which have no equivalents in other lan-
guages. To say that speakers of those other languages are deeply concerned
about some values which - "as it happens" - can only be formulated in
English means to give English a curiously privileged position in human-
kind's mental world. (To quote my colleague Cliff Goddard's ironic com-
ment on such methodological practices, "thank God for English!".)
The theory of cultural scripts rejects those practices, and seeks to for-
mulate norms, values and principles of language use in words which,
unlike impose or relevant, have equivalents in all other languages, that is
words which can be said to stand for universal human concepts. These
"universal words" (or word-like elements) are the same words in which
semantically encoded meanings can also be explicated.
The terminological distinction between "explications" and "cultural
scripts" can help clarify the boundary between those aspects of language
use which are semantically encoded and those which are not. Not every-
thing is semantically encoded but everything can be described in universal
human concepts. For example, "pragmatic" meanings encoded in "dimin-
utives" like doggie and birdie, in interjections like wow! or gee!, or in
tautologies like }var is ~ can be explicated in those concepts; and cul-
tural norms which are not encoded in any particular expressions can be
articulated in those concepts as a culture's "cultural scripts".
The main point is that neither conceptual artefacts encoded in the
English language nor Anglo "cultural scripts" can be legitimately used as
analytical tools for the interpretation of language use throughout the
world. The use of the Natural Semantic Metalanguage can free us from
such ethnocentrism and enable us to capture, in every case, the cultural
insider's point of view, while at the same time making that point of view
intelligible to the outsider.
Both explications and "cultural scripts" seek to articulate, in a rigor-
ous yet intelligible way, shared cultural representations. To quote from
Enfield (2000):
xx Introduction to the second edition
The very idea of the English language is a cultural and metalinguistic artefact.
So when we work with categories like English or Lao, this must be kept in
mind. And the same goes for 'Anglo' or 'Lao' culture. What we are really
talking about is some set of cultural representations - private representations
which are carried, assumed-to-be-carried and assumed-to-be-assumed-to-be-
carried - among some carrier group.... if we really want to characterize what
cultural representations unite groups of people, we had better start with the
cultural representation in question, and ask what group of people are united
by their sharing it, rather than starting with some group ... and asking what
cultural representations are shared among members. (Enfield 2000: 57)
To "start with the cultural representations" we need to have a framework
within which such representations can be identified - "from a native
speaker's point of view" (cf. Geertz 1984) and yet through concepts acces-
sible to cultural outsiders as well. The NSM theory, with its set of empiri-
cally established universal human concepts, provides such a framework.
The search for universals is of course important, but it must go in the
right direction. This book is based on the assumption that what is univer-
sal are the conceptual building blocks which we find in a tangible form
in all languages, and not some putative principles of "natural logic",
"conversation" or "politeness".
It is important to point out to the reader of this second edition that
the NSM semantic theory has developed considerably since the publica-
tion of the first edition - largely as a result of the theoretical as well as
empirical input from Cliff Goddard. Goddard himself has commented
on this development as follows:
In the thirty years since the publication of Semantic Primitives in 1972, the
mode of operation of the NSM research program has been akin to that of so-
called "normal science" (cf. Kuhn 1970; Lakatos 1970, 1978). There has been
internal consensus on the hard core of fundamental goals and assumptions -
the quest to identify the indefinable semantic elements in natural language and
to use these as a basis for a "self-explanatory" system of meaning representa-
tion. On the other hand, a number of auxiliary hypotheses have been revised
or replaced in the light of empirical work and the "model NSM" has passed
through a series of progressive refinements and expansions. (Goddard 2002a,
vol. 2: 314)
The expansions mentioned in the last sentence include the development
of the theory of cultural scripts and the new field of "ethnopragmatics"
(Goddard 2002b, in press a, b, c, and forthcoming), and the "refine-
ments" - the enlarged set of the universal semantic primes (roughly, the
double of that outlined in the first edition) and the construction of a
more or less complete model of universal grammar, presented in our re-
Introduction to the second edition XXI
cent edited book Meaning and Universal Grammar (Goddard and Wierz-
bicka 2002). If the present book were to be rewritten in the light of these
developments, the formulae included in it would be refined.
Since the new expanded set of universal human concepts constitutes
the major outcome of the NSM research over the last decade or so and
may be of interest to the reader of this second edition, I will include the
current set in Chapter 1, alongside with the 1991 version.
The doubling of the inventory of universal semantic primes must of
course be seen not only as a "refinement" but also as a major develop-
ment. In his insightful and generally very positive review of the 1991
Cross-Cultural Pragmatics James Matisoff (1996) has expressed some
scepticism with regard to the explanatory power of an inventory of only
27 elements, as it was at the time. The doubling of this set in more recent
NSM work vindicates Matisoff's scepticism. At the same time, I would
like to point out that most of the "new" post-1991 set of primes belong
to semantic domains which are less relevant to cross-cultural pragmatics
than the old ones, and also, that the actual analyses in the 1991 edition
of this book rely on more than 27 elements, although those additional
elements were regarded at the time as semantic "molecules" rather than
as semantic "atoms". Among the new primes which are relevant to many
"cultural scripts", the most important no doubt is TRUE (cf. e. g. Wierz-
bicka in press). In any case, I would encourage all those interested in
adopting the NSM framework for their own work on cross-cultural prag-
matics or indeed on any other aspect of language and culture to consult
also our 2002 edited book Meaning and Universal Grammar (Goddard
and Wierzbicka 2002).
In an article entitled "Cross-Cultural Literacy: A National Priority",
Luce and Smith (1987) wrote:
"Cross-cultural literacy" means that our citizenry knows how culture influ-
ences perceptions and actions. It no longer accepts cultural stereotypes and
cliches about other nations. It recognizes that American culture takes its place
beside other national cultures as one contruct within the spectrum of human
societies. Most importantly, cross-cultural literacy requires that Americans
know how to read the cultural cues of other nations and decode their meaning.
Within this decade, cross-cultural communications skill will become increas-
ingly an indispensable tool for every citizen. Cross-cultural literacy must be a
priority on our national agenda as we approach the end of the decade of the
1980s and near the 21
st
century. (Luce and Smith 1987: 4)
If "cross-cultural literacy" was justly seen as a priority in 1987, it is all
the more so in the post-September-l1th world of 2003 - and not only as
XXll Introduction to the second edition
a priority for the national agenda of the United States but also in Europe
and in many other parts of the world. Cultural stereotypes and cliches
are indeed no longer acceptable, but a wide-spread cross-cultural literacy
n1ust be seen as more important a goal than ever. The NSM semantic
theory based on universal human concepts offers a framework within
which the "cultural scripts" of different nations and different "lingua-
cultures" can be effectively articulated, taught and explained.
Canberra, January 2003
Anna Wierzbicka
References
[The articles on "cultural scripts" are marked with an asterisk]
Ameka, Felix
1994 Areal conversational routines and cross-cultural communication in
a multilingual society. In: H. Piirschel (ed.), Intercultural Communi-
cation, 441-469. Bern: Peter Lang.
1992 Interjections: The universal yet neglected part of speech. Journal of
Pragmatics 18 (2/3): 101-118.
Attinasi, John and Paul Friedrich
1995 Dialogic breakthrough: Catalysis and synthesis in life-changing di-
alogue. In: Bruce Mannheim and Dennis Tedlock (eds.), The Dia-
logic Emergence of Culture, 33-53. Urbana: University of Illinois
Press.
Besemeres, Mary
1998 Language and self in cross-cultural autobiography: Eva Hoffman's
"Lost in Translation". Canadian Slavonic Papers 40: 3-4.
2002 Translating one's self: Language and seljhood in crosscultural autobi-
ography. Oxford: Peter Lang.
Blum-Kulka, Shoshana, J. House, and Gabriele Kasper (eds.)
1989 Cross-cultural Pragn1atics: Requests and Apologies. Norwood, NJ:
Ablex.
Brown, Penelope and Stephen Levinson
1978 Universals in language usage: politeness phenomena. In: Esther
Goody (ed.), Questions and Politeness: Strategies in Social Interac-
tion, 56- 31 O. Cambridge: Cambridge University Press.
1978 Politeness: Some Universals in Language Usage. Cambridge: Cam-
bridge University Press.
Carbaugh, Donal
1988 Talking American. Cultural Discourses on DONOHUE. Norwood,
New Jersey: Ablex Publishing Corporation.
Introduction to the second edition XXlll
Clyne, Michael
1994 Intercultural Communication at Work: Cultural values in discourse.
Cambridge: Cambridge University Press.
Dalziell, Rosamund (ed.)
2002 Selves Crossing Cultures: Autobiography and Globalization. Mel-
bourne: Australian Scholarly Publishing.
D' Andrade, Roy and Claudia Strauss
1992 Hun1an Motives and Cultural Models. Cambridge: Cambridge Uni-
versity Press.
Davis, Wayne A
1998 Implicature: Intention, Convention, and Principle in the Failure of
Gricean Theory. Cambridge: Cambridge University Press.
Darnell, Regna
1994 Comments on Wolf's "Perilous ideas: Race, culture and people."
Current Anthropology 35.1: 7-8.
Enfield, Nick J.
2000 The theory of cultural logic: how individuals combine social intelli-
gence with semiotics to create and maintain cultural meaning. Cul-
tural Dynamics. 12.1: 35-64.
Geertz, Clifford
1984 From the Native's Point of View: On the Nature of Anthropologi-
cal Understanding. In: Richard A. Shweder and Robert A. LeVine
(eds.), Culture Theory: Essays on Mind, Self, and Emotion. 123-
136. Cambridge: Cambridge University Press.
Goddard, Cliff
1992 Traditional Yankunytjatjara ways of speaking: A semantic perspec-
tive. Australian Journal of Linguistics. 12.1: 93 - 122.
1996 The social emotions of Malay (Bahasa Melayu). Ethos. 24.3:
426-464.
*1997 Cultural values and cultural scripts of Malay (Bahasa Melayu).
Journal of Praglnatics 27.2: 183-201.
*2000 Cultural scripts and communicative style in Malay (Bahasa Mel-
ayu). Anthropological Linguistics 42.1: 81 -106.
2001 Cultural semantics and intercultural communication. In: D. Kil-
lick, M. Perry and A. Phipps (eds.), Poetics and Praxis of Lan-
guages and Intercultural Con1munication, 33-44. (Proceedings of
the conference at Leeds Metropolitan University December 1999).
Glasgow, Scotland: University of Glasgow French and German
Publications.
2002a The on-going development of the NSM research program. In: C.
Goddard and A. Wierzbicka (eds.), Meaning and Universal Gram-
mar: Theory and En1pirical Findings. vol. 2: 301 - 322. Amsterdam:
John Benjamins.
XXIV Introduction to the second edition
2002b Ethnosyntax, ethnopragmatics, sign-functions, and culture. In: N.
J. Enfield (ed.), Ethnosyntax: Explorations in Grammar and Culture,
52-73. Oxford: Oxford University Press.
2002c Overcoming terminological ethnocentrism. lIAS Newsletter 27, 28.
Leiden, The Netherlands: International Institute for Asian Studies.
in press a An ethnopragmatic perspective on active metaphors. Journal 0.[
Pragmatics.
*in press b "Cultural Scripts": A new medium for ethnopragmatic instruction.
In: Michael Achard and Susanne Niemeier (eds.), Cognitive Lin-
guistics, Second Language Acquisition, and Foreign Language Teach-
ing. Berlin/ New York: Mouton de Gruyter.
*in press c Directive speech-acts in Malay: An ethnopragmatic perspective. In:
Christine Beal (ed.), Les Cahiers de Praxematique.
*Forthc. Introduction to Cliff Goddard (ed.), Ethnopragn1atics: Understand-
ing discourse in cultural context. Berlin/New York: Mouton de
Gruyter.
Goddard, Cliff and Anna Wierzbicka
1997 Discourse and Culture. In: Teun A. van Dijk (ed.), Discourse as
Social Interaction, 231 - 257. vol. II of Discourse: A Multidiscipli-
nary Introduction. London: Sage Publications.
Goddard, Cliff and Anna Wierzbicka (eds.)
2002 Meaning and Universal Grammar: Theory and Empirical Findings.
2 vols. Amsterdam: John Benjamins.
Gordon, David and George Lakoff
1975 Conversational Postulates. In: Peter Cole and Jerry L. Morgan
(eds.), 83-106. Syntax and Semantics 3: Speech Acts. New York:
Academic Press.
Grice, H. P.
1975 Logic and conversation. In: Peter Cole and Jerry L. Morgan (eds.),
Syntax and Semantics 3: Speech Acts, 41-58. New York: Aca-
demic Press.
Harkins, Jean
1994 Bridging TIvo Worlds: Aboriginal English and cross-cultural under-
standing. Brisbane: University of Queensland Press.
Harnish R. M.
1976 Logical form and implicature. In: T. G. Bever, J. J. Katz, and J.
Langendoen (eds.), An Integrated Theory of Linguistic Ability, 313-
392. New York: Thomas Crowell.
Hoffman, Eva
1989 Lost in Translation: A Life in a nelV language. London: Heinemann.
Holland, Dorothy and Naomi Quinn
1987 Cultural Models in Language and Thought. Cambridge: Cambridge
University Press.
Introduction to the second edition xxv
Ide, Sachiko
1989 Formal forms and discernment. Multilingua 8: 223-248.
Kasper, Gabriele and Shoshana Blum-Kulka
1993 Interlanguage Pragmatics. Oxford: Oxford University Press.
Katriel, Tamar
1986 Talking Straight: Dugri Speech in Israeli Sabra Culture. Cambridge:
Cambridge University Press.
Kim, Young Yun
2001 Becoming Intercultural: An Integrative Theory of Communication
and Cross-Cultural Adaptation. Thousand Oaks, CA: Sage Publica-
tions.
Kochman, Thomas
1981 Black and White Styles in Con.flict. Chicago: University of Chi-
cago Press.
Kramsch, Claire
1993 Context and Culture in Language Teaching. New York: Oxford Uni-
versity Press.
Kuhn, Thomas
1970 The Structure of Scientific Revolutions. 2nd ed. Chicago: Chicago
University Press.
Lakatos, Imre
1970 Falsification and the methodology of scientific research pro-
grammes. In: I. Lakatos and A. Musgrave (eds.), Criticism and the
Grolvth of Knowledge, 91-196. Cambridge: Cambridge University
Press. [Republished in J. Worrall and G. Currie (eds.), Imre La-
katos: Philosophical Papers, vol. 1: 8- 101. Cambridge: Cambridge
University Press.]
1978 Introduction: Science and pseudoscience. In: J. Worrall and G.
Currie (eds.), Imre Lakatos: Philosophical Papers, vol. 1: 1-7.
Calnbridge: Cambridge University Press.
Leech, Geoffrey
1983 Principles of Pragmatics. London: Longn1an.
Levinson, Stephen
1983 Pragnlatics. Cambridge: Cambridge University Press.
Luce, Louise Fiber and Elise C. Smith
1987 Cross-cultural literacy: a national priority. In: Luce and Smith
(eds.), Toward Internationalism, 3-10. Cambridge: Newbury House
Publishers.
Lutz, Catherine
1988 Unnatural Emotions: Everyday Sentinlents on a Micronesian Atoll
and their Challenge to Western Theory. Chicago: University of Chi-
cago Press.
Matisoff, James
1979 Blessings, Curses, Hopes, and Fears: Psycho-Ostensive Expressions
in Yiddish. Philadelphia: Institute for the Study of Human Issues.
XXVI Introduction to the second edition
Matsumoto, Yoshiko
1988 Reexamination of the universality of face: politeness phenomena in
Japanese. Journal of Pragmatics 12: 403-426.
Shweder, Richard
1991 Thinking Through Cultures: Expeditions in Cultural Psychology.
Cambridge, MA: Harvard University Press.
1987
1997a
*1994a
*1996a
*1996b
*1994b
Searle, John
1975 Indirect speech acts. In: Peter Cole and Jerry Morgan (eds.), Syntax
and Semantics 3: Speech acts, 59-82. New York: Academic Press.
Sohn, Ho-min
1983 Intercultural communication in cognitive values: Americans and
Koreans. Language and Linguistics (Seoul) 9: 93 - 136.
Sperber, Dan and Deirdre Wilson
1986 Relevance: Conzmunication and cognition. Cambridge, MA: Har-
vard University Press.
Strauss, Claudia and Naomi Quinn
1997 A Cognitive Theory of Cultural Meaning. Cambridge: Cambridge
University Press.
Tannen, Deborah
1986 That's Not What I Meant! How conversational style nzakes or breaks
relationships. New York: Ballantine.
Wallenstein, Immanuel
1994 Comments on Wolf's 'Perilous ideas: race, culture and people'.
Current Anthropology. 35.1: 9-10.
Wierzbicka, Anna
1985 Different cultures, different languages, different speech acts: Eng-
lish vs Polish. Journal of Pragmatics 9: 145-178.
Boys Will Be Boys: "Radical Semantics" vs. "Radical Pragmatics."
Language 63: 95-114.
"Cultural Scripts": a new approach to the study of cross-cultural
communication. In: Martin Piitz (ed.), Language Contact and Lan-
guage Conflict. 69-88. Amsterdam/Philadelphia: John Benjamins.
Emotion, language and 'cultural scripts'. In: Shinobu Kitayama
and Hazel Rose Markus (eds.), Enlotion and Culture: Enlpirical
studies oj'mutual influence. 130-198. Washington D.C.: American
Psychological Association.
Contrastive sociolinguistics and the theory of "cultural scripts". In:
Marlis Hellinger and Ulrich Ammon (eds.), Contrastive Sociolin-
guistics, 313 - 344. Berlin/New York: Mouton de Gruyter.
Japanese cultural scripts: cultural psychology and "cultural gram-
mar". Ethos 24.3: 527- 555.
Understanding Cultures through their Key Words: English, Russian,
Polish, German, Japanese. New York: Oxford University Press.
Introduction to the second edition XXVll
1997b The double life of a bilingual: a cross-cultural perspective. In:
Michael Bond (ed.), Working at the Interface of Cultures: Eighteen
Lives in Social Science, 113-25. London: Routledge.
*1998 Gern1an cultural scripts: public signs as a key to social attitudes
and cultural values. Discourse and Society 9.2: 241 - 282.
*2002 Australian cultural scripts - bloody revisited. Journal of Pragn1atics
34.9: 1167-1209.
*in press Russian cultural scripts. Ethos.
Forthc. The English Language: Meaning, History and Culture.
Wolf, Eric R.
1994 Perilous ideas: race, culture and people. Current Anthropology
35.1: 1-7.
AcknowledgelTIents
I would like to express my gratitude to colleagues who at different times
have discussed with me problems explored in this book, and who have
offered comments on the earlier versions of some of the chapters, and
in particular, Felix Ameka, Andrzej Bogus}awski, Cliff Goddard, Jean
Harkins, Igor Mel'cuk, and Tim Shopen. I am particularly grateful to
Jean Harkins, who worked as my research assistant, and who made
innumerable valuable suggestions, as well as providing expert and
thorough editorial assistance.
I would also like to thank Mrs. Ellalene Seymour for her expert and
patient typing of the successive drafts.
In addition to several completely new chapters and other new work,
this volume includes also some chapters which had their starting point in
some articles published earlier. Although the contents of these chapters
is largely new (in length alone, none of the older articles comes to much
more than one third of the corresponding chapter), I would like to thank
the publishers of those earlier articles for their permission to make use
of the material contained in them:
1985: "A semantic metalanguage for a cross-cultural comparison of
speech acts and speech genres", Language in Society 14:491-514; "Dif-
ferent cultures, different languages, different speech acts: Polish vs.
English", Journal of Pragmatics 9:145-178.
1986: "A semantic metalanguage for the description and comparison
of illocutionary meanings", Journal of Pragmatics 10:67-107; "Italian
reduplication: cross-cultural pragmatics and illocutionary semantics",
Linguistics 24:287-315; "Precision in vagueness: the semantics of
English 'approximatives"', Journal of Pragmatics 10:597-614; "The
semantics of quantitative particles in Polish and in English", in:
Boguslawski - Bojar, 175-189.
1987: "Boys will be boys: 'radical semantics' vs. 'radical pragmatics"',
Language 63:95-114 (by permission of the Linguistic Society of
America).
1990: "The semantics of interjections", Journal of Pragmatics 14
(special issue on interjections, ed. by F. Ameka).
Canberra, September 1990 Anna Wierzbicka
Contents
Introduction to the second edition v
Acknowledgements XXIX
Chapter 1 Introduction: semantics and pragmatics 1
1. Language as a tool of human interaction 1
2. Different cultures and different modes of interaction 2
3. Pragmatics - the study of human interaction 5
4. The natural semantic metalanguage 6
5. The need for a universal perspective on meaning 9
6. The uniqueness of every linguistic system 10
7. The problem of polysemy 11
8. Semantic equivalence vs. pragmatic equivalence 12
9. Universal grammatical patterns 14
10. Semantics versus pragmatics: different approaches 15
10.1. 'Complementarism' 16
10.2. 'Pragmaticism' 17
10.3. 'Semanticism' 18
10.4. A fourth approach: two pragmatics 18
11. Description of contents 20
Chapter 2 Different cultures, different languages,
different speech acts 25
1. Preliminary examples and discussion 27
2. Interpretive hypothesis 30
3. Case studies 31
3.1. Advice 31
3.2. Requests 32
3.3. Tags 37
3.4. Opinions 41
3.5. Exclamations 45
4. Cultural values reflected in speech acts 47
4.1. Lexical evidence 47
4.2. Objectivism as a cultural value 49
4.3. Cordiality as a cultural value 50
xxxii Contents
4.4. Courtesy as a ~ u t u r value 56
5. Theoretical implications 59
6. Practical implications 64
Chapter 3 Cross-cultural pragmatics and different cultural values 67
1. ' Self-assertion' 72
1.1. 'Self-assertion' in Japanese and in English 72
1.2. 'Self-assertion' in black and white American English 78
1.3. Spontaneity, autonomy, and tum-taking:
English vs. Japanese 80
1.4. 'Spontaneous self-assertion' vs. 'regulated
self-assertion' :
black English vs. white English vs. Japanese 82
1.5. 'Self-assertion' as personal display:
black English vs. white English 84
1.6. 'Self-assertion' and 'good interpersonal relations' 85
2. 'Directness' 88
2.1. American culture vs. Israeli culture 89
2.2. 'Indirectness' in Japanese 93
2.3. Greek culture and American culture 95
2.4. 'Indirectness' and 'dissimulation' in Javanese 100
3. Further illustrations: same labels, different values 104
3.1. 'Intimacy' 105
3.2. 'Closeness' 108
3.3. 'Informality' 111
3.4. 'Harmony' 113
3.5. 'Sincerity' 115
4. Different attitudes to emotions 121
4.1. Polish culture 121
4.2. Jewish culture 122
4.3. American black culture 123
4.4. Japanese culture 126
4.5. Javanese culture 128
5. Conclusion 129
Chapter 4 Describing conversational routines 131
1. Conversational analysis:
linguistic or non-linguistic pragmatics? 131
Contents XXX111
2. 'Compliment response' routines 136
2.1. Upgrades 138
2.2. Contrastive opposites 139
2.3. Scaled-down agreements 140
2.4. Downgrades 140
2.5. Reassignment of praise 142
2.6. Returns 142
3. 'Compliment responses' in different cultures 143
4. Conclusion 147
Chapter 5 Speech acts and speech genres across
languages and cultures 149
1. A framework for analysing a culture's 'forms of talk' 149
1.1. The importance of folk labels 150
1.2. Two approaches 150
1.3. Some examples: English vs. Japanese 152
1.4. Another example: English vs. Walmatjari 158
1.5. The elimination of vicious circles 161
1.6. Evidence for the proposed formulae 161
1.7. The first-person format 162
1.8. The problem of other minds 164
2. Some Australian speech-act verbs 165
2.1. Chiack (chyack) 165
2.2. Yarn 170
2.3. Shout 173
2.4. Doh 177
2.5. Whinge 180
3. Some examples of complex speech genres 183
3.1. The black English dozens 183
3.2. The Hebrew 'dugri talk' 185
3.3. The Polish kawaf 188
3.4. The Polish podanie 192
4. Conclusion 196
Chapter 6 The semantics of illocutionary forces 197
1. Are illocutionary forces indeterminate? 197
1.1. Illocutionary forces as bundles of components 199
1.2. Illustration: the discrete and determinate
character of 'whimperatives' 202
XXXIV Contents
1.3. Syntax and illocutionary force 207
2. More whimperative constructions 211
2.1. Why don't you do X (tomorrow)? 211
2.2. Why do X? 213
2.3. How about X? 215
3. Additional remarks on the explication of
illocutionary forces 218
4. Selected conversational strategies 219
4.1. Tell you what, S! 219
4.2. Do you know, S? 220
4.3. Don't tell me S! 222
4.4. How many times have I told you (not) to do X! 223
4.5. Who's talking about doing X? 224
5. Tag questions 224
5.1. Tags with declarative sentences 224
5.2. Tags with imperative sentences 227
5.3. Why can't you (do X)! 229
5.4. OK? 231
6. Personal abuse or praise: You X! 232
7. Illocutionary forces of grammatical and
other categories 235
7.1. Modal verbs 235
7.2. Mental verbs 238
7.3. Particles and conjunctions 240
7.4. Interjections 243
7.5. Fixed expressions 245
7.6. Intonation 247
8. Comparing illocutionary forces across languages 248
9. Conclusion 252
Chapter 7 Italian reduplication: its meaning and
~ cultural significance 255
1. Italian reduplication: preliminary discussion 255
2. Discourse and illocutionary grammar 258
3. The illocutionary force of clausal repetition 260
4. The illocutionary force of Italian reduplication 263
5. Clausal repetition as a means of 'intensification' 268
6. The absolute superlative in Italian and in English 270
7. Illocutionary grammar and cultural style 276
8. Conclusion 282
Contents xxxv
Chapter 8 Interjections across cultures 285
1. Preliminary discussion 285
1.1. Interjections: physis and thesis
('nature' and 'convention') 285
1.2. Defining the concept of 'interjection' 290
1.3. Types of interjections 291
2. Volitive interjections 292
2.1. Interjections directed at animals 292
2.2. Interjections directed at people 293
2.2.1. The 'I want silence' group 293
2.2.2. The 'I don't want you in this place' group 296
2.2.3. The 'I want you to jump' group 298
2.2.4. The 'urging' group 298
2.2.5. The 'communication over distance' group 300
3. Emotive interjections 302
3.1. Interjections of 'disgust' and similar feelings 302
3.1.1. The Polish fu and the English yuk 302
3.1.2. The Russian fu 304
3.1.3. The Polish fe 306
3.1.4. The Yiddish feh 308
3.1.5. The Polish tfu and the Russian t'fu 310
3.1.6. 'Disgust' and bodily gestures 313
3.1.7. 'Disgust' and sound symbolism 315
3.2. 'General purpose' interjections 317
3.2.1. The Polish oj 318
3.2.2. The Russian oj 322
3.2.3. Ochs and achs 323
4. Cognitive interjections 326
4.1. The Polish aha and Russian aga 326
4.2. The Polish oho 331
4.3. The Polish 0 333
4.4. The English oh-oh 334
4.5. The Russian ogo 334
5. Conclusion 337
Chapter 9 Particles and illocutionary meanings 341
1. English quantitative particles 345
1.1. Non-approximative particles:
only, merely and just 346
XXXVI Contents
1.1.1. Only 346
1.1.2. Merely 348
1.1.3. lust 350
1.2. English approximative particles 354
1.2.1. Around and about 355
1.2.2. Approximately 358
1.2.3. Roughly 360
1.2.4. Almost and nearly 361
2. English temporal particles 367
3. Polish temporal particles 371
3.1. lui and jeszcze 371
3.2. Dopiero 376
4. Polish quantitative particles 379
4.1. Non-approximative particles 379
4.1.1. Tylko 379
4.1.2. Ai 380
4.1.3. Zaledwie 381
4.1.4. Ledwie 382
4.2. Polish approximative particles 384
4.2.1. o malo nie 384
4.2.2. Niemal and prawie 385
4.2.3. Blisko 388
5. Conclusion 389
Chapter 10 Boys will be boys: even 'truisms' are
culture-specific 391
1. The meaning of tautologies 391
1.1. Gricean maxims: universal or language-specific? 392
1.2. Problems in interpreting implicatures 397
1.3. Context as an excuse for analytical failure 400
2. English nominal tautologies: semantic representations 403
2.1. 'Realism' in human affairs 404
2.2. Tolerance for human nature 405
2.3. Tolerance at 'special times' 408
2.4. The limits of tolerance 410
2.5. Seeing through superficial differences 411
2.6. Recognising an irreducible difference 413
2.7. Tautologies of value 414
2.8. Tautologies of obligation 419
Contents XXXVll
3. Some comparisons from Chinese and Japanese 423
3.1. Chinese concessive tautologies 423
3.2. 'Irreducible difference', Chinese style 426
3.3. Chinese tautologies of unreserved praise 427
3.4. Japanese tautologies of 'a matter of course' 429
3.5. Japanese tautologies of irrelevance 430
4. Verbal tautologies 431
4.1. Future events 431
4.2. The immutability of the past 434
5. Is there a semantic invariant? 439
6. The deceptive form of English tautological
constructions 444
7. The culture-specific content of tautological patterns 446
8. Conclusion 448
Chapter 11 Conclusion: semantics as a key to
cross-cultural pragmatics 453
Notes 457
Bibliography 461
Subject and name index 487
Index of words and phrases 497
Chapter 1
Introduction: semantics and praglllatics
The fate of the earth depends on cross-cultural communication.
Deborah Tannen (1986:30)
1. Language as a tool of human interaction
This book is devoted to the study of language as a tool of human interac-
tion. It investigates various kinds of meanings which can be conveyed
in language (not in one language, but in different languages of the
world) - meanings which involve the interaction between the speaker
and the hearer.
It could be argued, of course, that all meanings involve interaction
between the speaker and the hearer: whether we talk about colours, ani-
mals, children, love, the fate of the universe, or even pure mathematics,
we use language as a tool of social interaction.
In some sense this is true. Nonetheless, there are words which involve
directly the concepts of 'I' and 'you', and interaction between 'I' and
'you', and there are others which do not. Similarly, there are grammati-
cal categories, and grammatical constructions, which involve these
concepts directly, and there are others which do not. For example, the
English words blue and yellow make no reference to the speaker, the
addressee, or the relationship between them; on the other hand, words
such as darling, bastard, already, yuk, thanks, or goodbye do. Similarly,
grammatical categories such as singular and plural number (dog vs. dogs)
or masculine and feminine gender (for example, la lille 'girl' vs. Ie
garfon 'boy' in French) do not involve the speaker, the addressee, or
the relationship between them; whereas categories such as diminutives
(doggie vs. dog), augmentatives (for example, problemon, problemazo
'big problem' vs. problema 'problem' in Spanish) or honorifics (for
example, otaku 'esteemed house' vs. ie 'house' in Japanese) do.
2 Introduction: semantics and pragmatics
At the level of grammatical constructions, the choice between an
m p r ~ t v (a) and a so-called 'whimperative' (b):
a. Sign this.
b. Would you sign this.
involves directly the relationship between the speaker and the addressee,
whereas the choice between a relative clause (a) and a participial
construction (b) does not:
a. The boy who was sitting in front . ..
b. The boy sitting in front . ..
This book, then, deals with words, categories, constructions, and
linguistic routines which involve interpersonal interaction, that is, which
involve, more or less directly, you and me. It is a book about you and
me, and about the different modes of interaction between you and me,
and, more particularly, between me and you (that is, between the speaker
and the hearer); and about cultural values and cultural norms which
shape these different modes of interaction.
2. Different cultures and different modes of interaction
There are many different possible modes of interaction between you and
me, between me and you. They depend partly on what you and I feel
and want at any particular time; but they depend also on who you and I
are - both as individuals and as members of particular social, cultural,
and ethnic groups. For example, if you and I are Japanese our interaction
will be different than it would be if we were both Americans or Russians.
And if we were both Americans, the prevailing modes of our interaction
would probably depend on whether we were white or black, Jewish or
non-Jewish, and so on.
Consider, for example, a typical Australian utterance such as Silly old
bugger!, recently used in public, in front of the television cameras, by
the Australian Prime Minister, Mr. Bob Hawke, during a meet-the-public
session, when he was goaded by an old-age pensioner about high parlia-
mentary salaries. One has to know a good deal about Australian culture
and society (cf. Chapter 5) to interpret correctly the communicative value
Different cultures and different modes of interaction 3
of this remark. In particular, one has to understand the link between
the common use of 'b-words' such as bugger, bastard, and bloody (cf.
Baker 1966:201) and the core Australian values of 'roughness', 'anti-
sentimentality', 'sincerity' and so on (cf. Renwick 1980; Wierzbicka, to
appear, chap. 11).
Similarly, one has to appreciate the core Australian values of
'mateship', 'toughness', 'anti-verbosity', 'anti-emotionality' and so on,
to appreciate the attitudes expressed in characteristic Australian greeting
e'xchanges (Bowles 1986:37; cf. Chapter 4):
G'day, mate, owyagowin?
Nobbad. Owsyerself? (Or: earn complain.)
In some cases, culture-specific modes of interaction have their own
folk names (cf. Chapter 5). This is the case with Black English speech
events such as 'rapping' or 'sounding', which can be illustrated with the
following characteristic utterances (Kochman 1972):
Baby, you're fine enough to make me spend my rent money.
(A 'rap' from a man to a woman.)
Baby, I sho' dig your mellow action. (Another e/xample of 'rap-
ping' to a woman.)
Yo mama is so bowlegged, she looks like a bite out of a donut.
(A 'sound' from a schoolboy to another schoolboy.)
But this is not necessarily always the case. Consider, for example, the
following conversation, from a short story by the Jewish-American
writer, Bernard Malamud:
[When he knocked, the door was opened by a thin, asthmatic, grey-haired
woman, in felt slippers.]
'Yes?' she said, expecting nothing. She listened without listening. He
could have sworn he had seen her, too, before but knew it was an
illusion.
'Salzman-does he live here? Pinye Salzman,' he said, 'the matchmaker?'
She stared at him a long minute. 'Of course.'
He felt embarrassed. 'Is he in?'
'No.' Her mouth, though left open, offered nothing more.
'The matter is urgent. Can you tell me where his office is?'
'In the air.' She pointed upward.
'You mean he has no office?' Leo asked.
'In his socks.' (... )
4 Introduction: semantics and pragmatics
'Where is he?' he insisted. 'I've got to see your husband.'
At length she answered, 'So who knows where he is? Every time he
thinks a new thought he runs to a different place. Go home, he will
find you.'
'Tell him Leo Finkle.'
She gave no sign she had heard.
(Malamud 1958:210-211)
The story is written in English, and it includes no unusual or non-
standard words, but the ways of speaking and of interacting reflected
here are those characteristic of Yiddish, not of (mainstream) American
English. Note in particular the use of No and Of course, the bare
imperatives Tell him and Go home, the rhetorical question Who knows?,
the irony, the wry humour, the bluntness and the gruffness (for discus-
sion, see Chapter 3 below).
And one last group of examples - English translations of typical
Yiddish blessings and curses (Matisoff 1979):
A lament to you, are you crazy or just feeble-minded?
Oh, you should be healthy, what a mess you've made here!
May he live - but not long.
A black year on her, all day long she chewed my ear off with
trivia!
My wife - must she live? - gave it away to him for nothing.
His son-in-law - may he grow like an onion with his head in
the earth - sold it to me.
Maybe my mother-in-law is going to visit us the day after
tomorrow, may the evil hour not come!
All such utterances encode important interactional meanings. This
book explores such meanings, and their cultural significance, and offers
a framework within which they can be described in an illuminating
and rigorous way.
Pragmatics - the study of human interaction 5
3. Pragmatics - the study of human interaction
The discipline studying linguistic interaction between 'I' and 'you' is
called pragmatics, and the present book is a work in pragmatics. It
differs, however, from other works in pragmatics in so far as it is also a
work in semantics - not in the sense that some chapters of the book
are devoted to pragmatics, and others, to semantics, but in the sense that
pragmatics is approached here as a part, or an aspect, of semantics; and
this is the major theoretical novum of the present approach.
I will explain what I mean by means of an example. Let us consider
first the words question and ask, sentences (questions) such as What time
is it?, so-called 'indirect questions' such as I don't know what time it is,
and so-called pre-questions, such as Do you know what time it is?
Traditionally, the word question would be described in a dictionary,
the sentence type illustrated by What time is it? would be discussed in a
chapter of a grammar devoted to 'interrogative constructions', and the
type illustrated by I don't know what time it is in a chapter of a grammar
devoted to 'indirect questions', whereas expressions such as Do you
know, Did you know or You know would be discussed (if at all) in some
works on 'discourse strategies', 'discourse markers', or on 'organisation
of conversation'. Thus, these different descriptions of words, grammati-
cal constructions, and 'pragmatic devices' would be discussed in totally
different types of works, and in totally different frameworks - as if
they had nothing in common whatsoever.
In fact, however, they are of course closely related. They all involve
crucially the concepts of 'knowing', 'not knowing', and 'saying'; and
they all involve the concepts of 'you' and 'I'. They all involve some
semantic components such as 'I don't know' or 'you don't know', 'I
say' and 'I want you to say', 'I want to know' or 'I want you to know'.
All these are 'interactional' (or 'pragmatic') meanings. To understand
human interaction we have to understand 'interactional' meanings
expressed in speech; and we have to have suitable analytical tools for
identifying and describing such meanings.
In the past, analytical tools of this kind were sorely lacking. Quite
apart from the compartmentalisation of linguistic descriptions, which
made it impossible to even raise the question of the semantics of human
interaction, there were simply no adequate tools for describing any
kind of interactional meanings. Standard lexicographic descriptions of
6 Introduction: semantics and pragmatics
words such as question or ask illustrate rather well the general level of
precision and clarity prevailing in the description of such meanings. For
example, Longman's ambitious Dictionary of the English Language
(LDOTEL 1984), which, according to its jacket blurb, "provides unrival-
led access to contemporary English and the way it is used", offers us
the following definitions:
question
interrogative -
command
response
(to) respond
(to) reply
a command or interrogative expression used to
elicit information or a response
an interrogative utterance, a question
the act of commanding
an act of responding
to write or speak in reply
to respond in words or writing
All such explanations of interactional meanings (like, incidentally, those
of any other meanings) are, clearly, totally circular. But it is an illusion
to think that circularity of this kind is exclusively a feature of diction-
aries (which are, after all, modest practical reference works), whereas
scholarly literature on language use is somehow different. It is not
different. It relies on various more or less technical-sounding labels
(such as, for example, 'face', 'distance', 'indirectness', 'solidarity', 'inti-
macy', 'formality', and so on), which are never defined; or if they are
defined, they are defined in ways which prove, sooner or later, to be just
as circular and obscure as traditional dictionary definitions. Furthermore,
they are defined in terms which are language-specific (usually, English-
specific), and which provide no language-independent, universal per-
spective on the meanings expressed in linguistic interaction.
4. The natural semantic metalanguage
To compare meanings one has to be able to state them. To state the
meaning of a word, an expression or a construction, one needs a seman-
tic metalanguage. To compare meanings expressed in different languages
and different cultures, one needs a semantic metalanguage independent,
in essence, of any particular language or culture - and yet accessible
and open to interpretation through any language.
The natural semantic metalanguage 7
1 propose for this purpose a 'natural semantic metalanguage', based on
a hypothetical system of universal semantic primitives, which my
colleagues and 1 have developed over more than two decades (see, in
particular, Boguslawski 1966, 1972, 1975, 1981a,b, 1989; Wierzbicka
1972, 1980, 1987, 1988, 1989a,b; Goddard 1989a,b); and this is the
metalanguage employed in the present book.
This means that 1 will try to state the meanings under consideration
in terms of simple and intuitively understandable sentences in natural
language. This, 1 believe, will ensure that the proposed semantic explica-
tions will be immediately verifiable and intuitively revealing. But the
subset of natural language in which the explications are formulated is
highly restricted, standardised, and to a large extent language-independ-
ent (that is, isomorphic to equivalent subsets of other natural languages).
For this reason, the natural language used in the explications - a kind
of highly reduced 'basic English ' - can be viewed as a formal semantic
metalanguage.
The metalanguage applied in the present work is, so to speak, carved
out of natural language - any natural language. For practical reasons,
the version of the metalanguage employed here is carved out of English,
but it could be just as easily carved out of Russian, Latin, Japanese, or
Swahili, because it is based, by and large, on what 1 believe to be the
universal core of natural languages. For example, if 1 say in an
explication: 'I want', 1 mean something that could be just as easily repre-
sented as 'ja xocu' (Russian) or 'ego volo' (Latin). The expression 'I
want' is used here, therefore, not as part of the 'normal' English
language, but as part of the English-based version of the universal
semantic metalanguage.
The metalanguage in question is a technical, artificial language, not a
natural language; nonetheless, it is appropriate and illuminating, 1 think,
to call it a 'natural semantic metalanguage', (cf. Goddard 1989a,b), be-
cause it is derived entirely from natural language and because it can be
understood via natural language without any additional arbitrary signs
and conventions. Arbitrary signs and conventions are not allowed in this
metalanguage, because their meaning would have to be explained - and
these explanations, in their turn, would not be intelligible unless they
were couched in immediately understandable natural language. (On the
other hand, it is allowed to use 'iconic' conventions, such as spatial
arrangement of components, the use of separate lines for different chunks
of meanings, and the like.)
8 Introduction: semantics and pragmatics
Adjectives
good
bad
Addendum to Chapter 1
The 1991 version of the Natural Semantic Metalanguage included only
27 hypothetical semantic primes, although, as mentioned earlier, a
number of auxiliary concepts, regarded at the time as semantic "mole-
cules" rather than "atoms" were also used in the explications. The cur-
rent, greatly expanded, set includes 60 or so elements. Both versions, the
1991 and the current one, are given below. *
Universal semantic primes (1991 version)
Pronouns Determiners Classifiers
I this kind of
you the same part of
someone two
something all
Linkers
like
because
Place/Time
place
time
after (before)
above (under)
Modals
can
if/imagine
Verbs
want
don't want
say
think
know
do
happen
Universal semantic primes (2003 version)
Substantives: I, YOU, SOMEONE, PEOPLE, SOMETHING/THING, BODY
Determiners: THIS, THE SAME, OTHER
Quantifiers: ONE, TWO, SOME, ALL, MUCH/MANY
Evaluators: GOOD, BAD
Descriptors: BIG, SMALL
Mental predicates: THINK, KNOW, WANT, FEEL, SEE, HEAR
Speech: SAY, WORDS, TRUE
Actions, events and movement: DO, HAPPEN, MOVE
Existence and possession: THERE IS, HAVE
Life and death: LIVE, DIE
Time: WHEN/TIME, NOW, BEFORE, AFTER, A LONG TIME, A SHORT TIME, FOR SOME
TIME, MOMENT
Space: WHERE/PLACE, HERE, ABOVE, BELOW, FAR, NEAR, SIDE, INSIDE, TOUCHING
(CONTACT)
Logical concepts: NOT, MAYBE, CAN, BECAUSE, IF
Intensifier, augmentor: VERY, MORE
Taxonomy, partonomy: KIND OF, PART OF
Similarity: LIKE
* This paragraph replaces an outdated paragraph from the 1991 edition. This is the only
paragraph in the text of the first edition replaced with a new version. The only other change
in the 1991 text is the addition of the 2003 table of universal semantic primes, on the
same page.
The need for a universal perspective on meaning 9
5. The need for a universal perspective on meaning
It is impossible for a human being to study anything - be it cultures,
language, animals, or stones - from a totally extra-cultural point of
view. As scholars, we remain within a certain culture, and we are inevi-
tably guided by certain principles and certain ideals which we know are
not necessarily shared by the entire human race.
We must also rely on certain initial concepts; we cannot start our
inquiry in a complete conceptual vacuum. It is important, however, that
as our inquiry proceeds, we try to distinguish what in our conceptual
apparatus is determined by the specific features of the culture to which
we happen to belong, and what can be, with some justification, regarded
as simply human.
Trying to explore both the universal and the culture-specific aspects
of meaning, we should beware of using concepts provided by our own
culture or by our own scholarly tradition as culture-free analytical tools
(cf. Lutz 1985). As human beings, we cannot place ourselves outside all
cultures. This does not mean, however, that if we want to study cultures
other than our own all we can do is to describe them through the prism of
our own culture, and therefore to distort them. We can find a point of
view which is universal and culture-independent; but we must look for
such a point of view not outside all human cultures, (because we cannot
place ourselves outside them), but within our own culture, or within any
other culture that we are intimately familiar with.
To achieve this, we must learn to separate within a culture its idiosyn-
cratic aspects from its universal aspects. We must learn to find 'human
nature' within every particular culture. This is necessary not only for the
purpose of studying 'human nature' but also for the purpose of studying
the idiosyncractic aspects of any culture that we may be interested in. To
study different cultures in their culture-specific features we need a
universal perspective; and we need a culture-independent analytical
framework. We can find such a framework in universal human concepts,
that is in concepts which are inherent in any human language.
If we proceed in this way, we can study any human culture without the
danger of distorting it by applying to it a framework alien to it; and we
can aim both at describing it 'truthfully', and at understanding it.
We cannot understand a distant culture 'in its own terms' without
understanding it at the same time in our own terms. What we need for
10 Introduction: semantics and pragmatics
real 'human understanding' is to find terms which would be both 'theirs'
and 'ours'. We need to find shared terms; that is, universal concepts. I
suggest that we can find such concepts in the 'universal alphabet of
human thoughts' (Leibniz 1903:430), that is, in the indefinable (Le.
semantically simple) words and morphemes of natural language, (such
as I, you, someone, something, this, think, say, want, or do), which can
be found, it seems, in all the languages of the world.
6. The uniqueness of every linguistic system
Every language is a self-contained system and, in a sense, no words or
constructions of one language can have absolute equivalents in another.
The idea that there might be some linguistic elements which are univer-
sal in the sense of having absolute equivalents in all the languages of
the world is of course all the more fanciful.
However, as soon as we abandon the notion of absolute equivalents
and absolute universals, we are free to investigate the idea of partial
equivalents and partial universals; and if the former notion is sterile and
useless, the latter idea is fruitful and necessary.
What I mean by 'partial universals' is this. Within a particular
language, every element belongs to a unique network of elements, and
occupies a particular place in a unique network of relationships. When
we compare two, or more, languages we cannot expect to find identical
networks of relationships. We can, nonetheless, expect to find certain
correspondences.
To put it differently, although every language has its own unique
structure and its own unique lexicon (embodying unique semantic
configurations), nonetheless there are certain areas of languages which
can be regarded as mutually isomorphic (some examples are given in
the sections which follow). It is this (limited) isomorphism in grammar
and in the lexicon that gives sense to the notion of semantic universals.
The metalanguage employed in the present book is based on such
putative universals.
The problem of polysemy 11
7. The problem of polysemy
The search for lexical universals may seem to be a purely empirical
task: laborious, to be sure, but relatively straightforward. In fact, how-
ever, the presence or absence of a word for a given concept cannot be
established by any mechanical, checklist method. The search is empiri-
cal, but it also has necessarily an analytical dimension. Above all, there
is the problem of polysemy. For example, 1 have postulated 'you' and 'I'
as universal semantic primitives, but what I mean by 'you' is 'you SG'
('thou'), rather than 'you PL' or 'you SGjPL'. Yet one doesn't have to
look further than modem English to find a language which doesn't seem
to have a word for 'thou'. To maintain the claim that 'thou' is a lexical
universal we would have to posit polysemy for the word you: (1) 'you
SG', (2) 'you PL'. Initially, this seems an unattractive solution, but I
think there are good reasons for accepting it. Polysemy is a fact of life,
and basic, everyday words are particularly likely to be polysemous (cf.
Zipf 1949). For example, say is polysemous between its abstract sense,
which ignores the physical medium of expression (for example What
did he say in his letter?, The fool said in his heart: there is no God), and
its more specific sense, which refers to oral speech only. Know is
polysemous between the two senses which are distinguished in French
as savoir and connaftre, and in German as wissen and kennen (cf. I know
that this is not true vs. I know this man).
It goes without saying that polysemy must never be postulated lightly,
and that it has always to be justified on language-internal grounds; but
to reject polysemy in a dogmatic and a priori fashion is just as foolish
as to postulate it without justification. In the case of the English word
you, I think its polysemy can be justified on the basis of the distinction
between the forms yourself and yourselves; the choice between yourself
and yourselves is determined by the choice between youse and you
pL
(cf. youse yourself vs. you
pL
yourselves).
There is nothing surprising in the fact that one word may have two
meanings, one indefinable and one definable. It is more surprising if one
word appears to have two different indefinable meanings. In fact, how-
ever, the evidence available so far suggests that there are no languages
in the world which would use the same word for 'you' and 'I'. More
generally, there appear to be no languages in the world which wouldn't
have special (separate) words for these two vital concepts.
12 Introduction: semantics and pragmatics
8. Semantic equivalence vs. pragmatic equivalence
If there are scholars who - like the ordinary monolingual person -
believe that most words in one language have exact semantic equivalents
in other languages, there are also those who believe that no words in one
language can have exact equivalents in many other languages, let alone
in all the languages of the world. For example, they say, there are
languages which have no personal pronouns, no words for 'you' or 'I'.
Japanese is sometimes cited as an example of this. In my view, this is a
fallacy. For cultural reasons, Japanese speakers try to avoid the use of
personal pronouns (cf. Bamlund 1975b; Suzuki 1986) and the language
has developed a wealth of devices that allow its speakers to avoid such
overt reference without producing any misunderstandings. For example,
there are certain verbs in Japanese (so-called honorific verbs) which are
never used with respect to the speaker; and there are 'humble', self-
deprecating verbs, which are never used with respect to the addressee;
the use of such verbs often sufficiently identifies the person spoken
about and the person addressed as to make an overt reference to 'you'
and 'I' unnecessary. But the words for 'you' and 'I' do exist, and can
be used when it is necessary or desired.
It is also true that many languages, especially Southeast Asian
languages, have developed a number of elaborate substitutes for 'you'
and 'I', and that in many circumstances it is more appropriate to use
some such substitute than the barest, the most basic pronoun. For
example, in a polite conversation in Thai, the use of the basic words for
'you' and 'I' would sound outrageously crude and inappropriate. Instead,
various self-deprecating expressions would be used for 'I', and various
deferential expressions for 'you'. Many of the expressions which stand
for 'I' refer to the speaker's hair, crown of the head, top of the head, and
the like, and many of the expressions which stand for 'you' refer to the
addressee's feet, soles of the feet, or even to the dust underneath his
feet, the idea being that the speaker is putting the most valued and
respected part of his own body, the head, at the same level as the lowest,
the least honourable part of the addressee's body (cf. Cooke 1968;
Palakomkul 1975). But this does not mean that Thai has no personal
pronouns, no basic words for 'you' and 'I'.
A language may not make a distinction which would correspond to
that between the words 'he' and 'she', and in fact many languages, for
example Turkish, have just one word for 'he' and 'she', undifferentiated
Semantic equivalence vs. pragmatic equivalence 13
for sex. But no known language fails to make a distinction between the
speaker and the addressee, i.e. between 'you' and 'I'.
This does not mean that the range of use of the words for 'you' and 'I'
is the same in all languages. For example, in Japanese, the word ore,
which Japanese English dictionaries gloss as 'I', has a range of use in-
comparably more narrow than the word I has in English. Thus, in a
recent study of the use of the first and second person pronouns
(Kurokawa 1972), it was found that none of the women in the sample
used ore, whereas 90% of the men did - along with boku (100%),
watashi (80%), watakushi (50%) and atashi (80%). It was also found
that "the pronoun ore 'I' is often used among male adult speakers only
in such very informal occasions as between two close friends and at
home. It is not an exaggeration to say that in many elementary schools
the use of this pronoun ore is discouraged by the teacher. ... This
pronoun is almost never introduced in texts for an elementary, or an
intermediate, Japanese course for English speaking students." (Kurokawa
1972:231). The survey also shows that "men use ore more frequently
when talking with their wives than when talking with their parents: 44%
versus 33%" (1972:232).
What does ore mean, then? It may be considered 'rude' for a child to
use ore to other children at school, but ore cannot mean 'I + disrespect',
because if it did it would not be permissible for a man to use it when
speaking to his parents. This suggests that ore means simply 'I' - and
that there are no invariant semantic components which could be always
attributed to it other than 'I'. The heavy restrictions on its use must
therefore be attributed to cultural rather than semantic factors. In a soci-
ety where references to oneself are in many situations expected to be
accompanied by expressions of humility or deference, a bare 'I' becomes
pragmatically marked, and it must be interpreted as either very intimate
or very rude. But this pragmatic markedness should not be confused with
demonstrable semantic complexity.
Above all, it should be pointed out that words such as the Japanese ore
'I' or kimi 'you' (or French tu, or German du), cannot be further defined
within the languages to which they belong. Even if someone insisted that
words of this kind can be defined via English, for example, along the
following lines:
ore - 'I; I don't have to show respect for you'
kimi (tu, du) - 'you; I don't have to show respect for you'
14 Introduction: semantics and pragmatics
explications of this kind could not be translated into Japanese, French, or
German without regressus ad infinitum (for what words would be used
for you and I in the explication?). We have to conclude, therefore, that
words of this kind are true semantic primitives of the languages in
question. To say that they are not semantic primitives, but that their
inherent complexity can be shown only via definitions phrased in
English, not in the languages to which they belong, would be a case of
blatant ethnocentrism. Since, however, these primitives (of the Japanese,
French, or German language) can be matched semantically across
language boundaries, we can acknowledge their analogous (indefinable)
position within the language systems to which they belong by calling
them universal semantic primitives, and by equating them in semantic
explications - despite the huge cultural differences reflected in their
different frequency and different range of use.
9. Universal grammatical patterns
But if the supposed lexical universals are embedded, in each language,
in language-specific grammatical patterns, can they really be matched
and identified cross-linguistically? In any case, words or morphemes by
themselves cannot really express any meanings: they can only contribute
in a certain way to the meaning expressed by a sentence. If we want to
identify meanings cross-linguistically we must look not for isolated
lexical items but for commensurable lexical items used in commensu-
rable sentences. This means that we must look not only for commensu-
rable lexical items but also for commensurable grammatical patterns.
It seems clear that the great majority of grammatical patterns of any
given language are language-specific. It is possible, however, that there
are also some patterns which are universal. In fact, if cross-cultural
understanding is possible at all, despite the colossal variation in language
structures, there must be some common core of 'human understanding',
and this common core must rely not only on some shared or matching
lexical items but also on some shared, or matching, grammatical patterns
in which those shared lexical items can be used.
To put it differently, there must be some 'atomic sentences' (cf.
Russell 1962), or 'kernel sentences' (cf. Chomsky 1957), which can be
said in any language, and which can be matched across language bounda-
Semantics versus pragmatics: different approaches 15
ries. The grammar of these 'atomic sentences' must consist in the
possible distribution patterns of the 'atomic elements' (that is, the lexical
indefinables). Trying to discover those patterns we should look, there-
fore, at the lexical indefinables themselves, and try to see what their
possibilities of co-occurrence might be. In searching for universal
grammatical patterns, therefore, we should not look for any universals
of form; rather, we should look for universals of combinability.
The search for such simple and 'language-independent' grammatical
patterns has begun fairly recently and is still in its early stages (cf.
Wierzbicka 1988, in press a, b). The explications proposed in the present
work employ a kind of reduced (English) syntax which is relatively
simple and relatively language-independent, without being simple or
universal in any absolute sense. Above all, I try to rely on simple clauses
rather than on complex sentences and to avoid participial constructions,
relative clauses, nominalisations, and other similar pieces of complex,
language-specific syntactic machinery. I do not, however, try in this
work to go as far as possible in the direction of simplicity and universal-
ity, because this would often increase the length of explications and
make them more difficult to read. I aim at a compromise between
simplicity and universality on the one hand and the reader's convenience
on the other.
10. Semantics versus pragmatics: different approaches
Leech (1983:6) distinguishes three different ways of viewing the rela-
tionship between semantics and pragmatics, which he summarises
usefully in the form of three diagrams, shown in Figure 1. He labels
the three approaches 'semanticism' (A), 'complementarism' (B), and
'pragmaticism' (C).
The classical Morrisian (1938) position, which divides the study of
sign systems into syntax, semantics and pragmatics, is an instance of
'complementarism'. The philosophical tradition in the study of language
which started with Wittgenstein's Philosophical Investigations (1953)
and which urged, 'Don't ask for meaning, ask for use', is an instance of
'pragmaticism'. The 'generative semantics' of the early 1970s, which
tried to present the illocutionary force of an utterance as part of its
semantic structure, can be said to have represented 'semanticism' (see
16 Introduction: semantics and pragmatics
A
semantics
r----------,
I (pragmatics) I
L .J
B
semantics
pragmatics
c
r---------,
I (semantics) I
L -J
pragmatics
Figure 1. Three views of the relationship between semantics and pragmatics
the articles in Cole - Morgan 1975). All three of these approaches
present serious difficulties, which will be discussed briefly below.
10.1. 'Complementarism'
Morris wanted to separate the relations between signs and 'reality' from
the relations between signs and their users. But the very nature of natural
language is such that it doesn't separate extralinguistic reality from the
psychological and social world of language users.
Language is an integrated system, where everything 'conspires' to
convey meaning: words, grammatical constructions and various 'illocu-
tionary' devices (including intonation). Accordingly, one might argue
that linguistics falls naturally into three parts, which could be called
lexical semantics, grammatical semantics, and illocutionary semantics. A
Morrisian division of the study of signs into semantics, syntax, and
pragmatics may make good sense with respect to some artificial sign
systems, but it makes no sense with respect to natural language, whose
syntactic and morphological devices (as well as illocutionary devices)
are themselves carriers of meaning. In natural language, meaning con-
sists in human interpretation of the world. It is subjective, it is anthropo-
centric, it reflects predominant cultural concerns and culture-specific
modes of social interaction as much as any objective features of the
Semantics versus pragmatics: different approaches 17
world 'as such'. 'Pragmatic (attitudinal) meanings' are inextricably
intertwined in natural languages with meanings based on 'denotational
conditions' (see for example Wierzbicka 1980, 1987; see also Padu-
ceva 1985).
Since the meanings conveyed in natural language are inherently
subjective and anthropocentric, they cannot be neatly divided into 'refer-
ential' and 'pragmatic', or 'denotational' and 'attitudinal'. What is
needed, therefore, is a unified semantic framework, equally suitable for
describing the meaning of 'cultural kinds' (such as cup and mug in
English, or sake in Japanese), 'natural kinds' (such as cat and dog in
English, or nezumi 'rat/mouse' in Japanese), interactional verbs (such as
promise, vow, or pledge in English, or materit'sja 'mother-swear' in
Russian), and so on. All such meanings are culture-specific, subjective,
and anthropocentric (see Wierzbicka 1985a,b, 1987), 'referential' and
'pragmatic' at the same time. For example, Leech's 'complementarist'
position forces him to analyse illocutionary forces such as requesting,
promising, and ordering under 'pragmatics', and the meaning of verbs
such as request, promise, and order, under 'semantics', as if the two
tasks had nothing in common, and as if the so-called illocutionary force
of requesting, promising, or ordering wasn't simply a function of the
English verbs request, promise, and order.
10.2. 'Pragmaticism'
The approach that Leech has called 'pragmaticism' has perhaps more to
offer, because it creates no artificial gulf between 'pragmatic meanings'
and 'denotational meanings' and recognises the anthropocentric nature
of natural language, where 'man' (the language user) is truly a measure
of all things, and where 'objective' aspects of meaning are inextricably
linked with 'subjective' and interactional ones.
Yet 'pragmaticism', too, proves very hard to apply fruitfully when it
comes to actual description of meanings, especially in a cross-cultural
perspective, because it has no rigorous framework for description and
comparison, no firm grid in terms of which the endless vagaries of
language use can be rigorously analysed and interpreted.
To try to describe language use without such a grid is like trying to
describe phonological systems of different languages without having a
universal phonetic alphabet of any sort. Not surprisingly, many linguists
accustomed to high standards of rigour in domains such as phonology,
18 Introduction: semantics and pragmatics
syntax, or historical linguistics reject linguistic articles and books based
on the philosophy of 'pragmaticism' as 'woolly', 'waffly', and arbitrary.
10.3. 'Semanticism'
In the present writer's view, the approach which Leech calls 'semanti-
cism' has much more to offer to the study of meaning in natural
language, because it can provide it with a firm basis and can allow it
to combine insight with rigour. Natural language is a system for convey-
ing meaning, and any integration of linguistic science can be achieved
only on the basis of meaning.
The fact that a well-known linguistic school which advocated a 'radi-
cally semantic' position (' generative semantics ') has failed, and has
acknowledged its defeat (see Newmeyer 1980:167-173; Lakoff 1986:
584-585), doesn't mean that there is something inherently wrong with a
'radically semantic' orientation as such. One cannot describe and com-
pare meanings in a non-arbitrary way without a well-justified set of
(candidates for) universal semantic primitives. Generative semanticists
didn't strive to discover such a set (although they did like to refer, in the
abstract, to some unidentified 'atomic predicates '). One can argue that
this was the main cause of their failure (in pragmatics, and in semantics
in general), not their 'radically semantic' approach. What they lacked
was a methodology which would lend coherence and unity to the field of
semantics, and which would define a well-justified boundary around it.
Linguistic semantics and linguistic pragmatics are one. What applies
to colour semantics, kinship semantics, speech-act semantics, to the
semantics of natural kinds, cultural kinds, emotions, and so on applies
also to the semantics of interpersonal attitudes.
10.4. A fourth approach: two pragmatics
But can all aspects of pragmatics be handled by means of a universal
semantic framework, the same framework which can also be used for all
other areas of meaning?
Probably nobody would want to go so far as to claim that. The term
'pragmatics' has been applied to a very wide and heterogeneous range
of phenomena, including 'conversational analysis', 'linguistic etiquette',
'acquisition of communicative competence', and so on. In fact, many
Semantics versus pragmatics: different approaches 19
scholars have suggested that 'pragmatics' is no more than a wastepaper
basket, where everything that has to do with language but which cannot
be treated rigorously is thrown. This position gives 'pragmatics' a very
broad scope indeed, but it leaves the 'core linguistics' greatly impover-
ished and deprived of a component which is essential to a coherent and
integrated description of linguistic competence.
In my view, the only possible solution to this dilemma is to recognise
that there are two pragmatics, differing from one another not so much in
subject matter as in methodology. There is a linguistic pragmatics, which
can form a part of a coherent, integrated description of linguistic com-
petence, and there is another pragmatics, or other pragmatics (in the
plural): a domain or domains of the sociologist, the psychologist, the
ethnomethodologist, the literary scholar, and so on.
As Hugo Schuchardt (1972:67) pointed out, the unity of a scholarly
discipline is created by its coherent methodology, not by any inherent
unity of the subject matter. Pragmatics is, up to a point, an integral
part of linguistics, and the boundary between linguistic pragmatics
and nonlinguistic pragmatics is determined by the stretching capacities
of a coherent unified linguistic framework.
Attitudinal meanings can be treated in the same descriptive framework
as any other kinds of meaning. They can therefore be regarded as belong-
ing to semantics and, ipso facto, to 'core' linguistics. There is no gulf
between linguistic pragmatics and linguistic semantics; on the contrary,
linguistic pragmatics can be fruitfully seen as part of linguistic seman-
tics. But there is a gulf between linguistic pragmatics and various other,
heterogeneous, considerations of language use. This leads us to propose
a fourth diagram, shown in Figure 2, in addition to the three proposed
by Leech.
This diagram represents a 'radically semantic' approach to meaning,
with so-called 'pragmatic meanings' being treated in exactly the same
way, and being described in exactly the same framework, as any other
kind of meaning. But this doesn't mean that anything that has ever been
called 'pragmatics' could, or should, be swallowed by semantics.
20 Introduction: semantics and pragmatics
j:::)
I'Q
r----------7------t/)Q
Semantics tics
Non
pragmatics

Figure 2. A 'radically semantic' approach to meaning
11. Description of contents
Chapter 2, 'Different cultures, different languages, different speech
acts', discusses a number of differences between two languages, English
and Polish, in the area of speech acts, and links these differences with
different cultural norms and cultural assumptions. It is shown that
English, as compared with Polish, places heavy restrictions on the use of
the imperative and makes extensive use of interrogative and conditional
forms. Features of English which have been claimed to be due to univer-
sal principles of politeness are shown to be language-specific and due to
specific cultural norms and cultural traditions. Linguistic differences
are shown to be associated with cultural values such as individualism
and respect for personal autonomy in the case of English, and cordiality
in the case of Polish. Furthermore, certain characteristic features of
Australian English are discussed and illustrated, and are shown to reflect
Description of contents 21
some features of the Australian national ethos. Implications for the
theory of speech acts and for intercultural communication are discussed.
In particular, certain influential theories of speech acts, based largely on
English (in particular, Searle's theory) are shown to be ethnocentric
and dangerous in their potential social effects.
Chapter 3, 'Cross-cultural pragmatics and different cultural values',
uses a much wider range of examples (in particular, from Japanese,
Black American English, Yiddish, and Hebrew), to show that differences
in the ways of speaking associated with different languages are profound
and systematic, and that they reflect, and can be explained in terms of,
independently established differences in cultural traditions, cultural
values, and cultural priorities. It demonstrates the anglocentrism of
supposedly universal 'maxims' of human conversational behaviour of
the kind put forward by Grice (1975) or Leech (1983). It also shows how
progress in cross-cultural pragmatics has been hampered by the use of
inadequate conceptual tools: in particular, of unanalysed, obscure and
protean global labels such as 'directness', 'self-assertion', 'distance',
'intimacy', 'solidarity', 'harmony', 'informality', and so on, which have
led to paradoxical and contradictory conclusions; and it proposes a
method whereby different communicative styles can be clarified in
terms of 'cultural scripts' written in the metalanguage of universal
semantic primitives.
Chapter 4, 'Describing conversational routines', shows that while
considerable effort has gone into the description and comparison of con-
versational routines associated with different languages and different
cultures, much less has been achieved in this important area than might
have been - because not enough thought has been given to the vital
question of a standardised and 'culture-free' metalanguage in which such
comparisons could be fruitfully carried out. To show how the use of the
natural semantic metalanguage can facilitate this task I examine, in
particular, a number of generalisations suggested in Pomerantz's (1978)
paper 'Responses to compliments', and I show how these generalisations
could be reformulated to make them both clear and verifiable. I also
examine a number of other conversational routines, trying to show how
the use of the natural semantic metalanguage can bring a new level of
rigour to conversational analysis, and can free it from ethnocentric bias.
Chapter 5, 'Speech acts and speech genres across languages and
cultures', discusses a number of speech acts and speech genres from
English, Polish, Japanese, Hebrew, and Walmatjari (an Australian
Aboriginal language), approaching them through the words which name
22 Introduction: semantics and pragmatics
them (that is to say, through their folk labels). It is claimed that folk
names of speech acts and speech genres provide an important source of
insight into the communicative styles most characteristic of a given
society, and reflect salient features of the culture associated with a given
language; and that to fully exploit this source one must carry out a
rigorous semantic analysis of such names, and express the results in a
culture-independent semantic metalanguage. This is shown in detail
through the semantic analysis of a group of Australian English speech-
act verbs, together with a discussion of traditional Australian values
and the Australian national ethos.
Chapter 6, 'The semantics of illocutionary forces', examines a wide
range of English constructions and expressions encoding certain
modes of interpersonal interaction, and spells out their meaning (or
their 'illocutionary force'). For example, different types of tag questions
and different types of 'interrogative directives' ('whimperatives ') are
discussed, and both the similarities and differences between them are
made explicit. Here as in the other chapters of the book, the analysis
takes the form of decomposition of illocutionary forces into their com-
ponents, which are formulated in the natural semantic metalanguage. It
is argued that the decomposition of illocutionary forces illustrated in
this chapter offers a safe path between the Scylla of the 'performative
hypothesis' (which has proved to be empiricially inadequate and theo-
retically unjustifiable) and the Charybdis of the 'autonomous grammar',
which tries to divorce the study of language structure from the study
of language use.
Chapter 7, 'Italian reduplication: its meaning and its cultural signifi-
cance', constitutes a case study of one culture-specific pragmatic device:
the Italian 'reduplication' (for example bella bella 'beautiful beautiful '),
examined against the background of various other 'intensification de-
vices', such as, for example, the absolute superlative (for example I am
most grateful). It is demonstrated that subtle pragmatic meanings such
as those conveyed in Italian reduplication can be identified and distin-
guished from other, related meanings if ad hoc impressionistic comments
are replaced with rigorous semantic explications; and it is shown how
a semantic metalanguage derived from natural language can be used
for that purpose. It is also argued that syntactic reduplication belongs
to a system of pragmatic devices which reflect, jointly, some charac-
teristic features of Italian communicative style. More generally, it is
argued that illocutionary grammar can be linked directly with 'cultural
style', and that cross-cultural pragmatics can gain considerably in both
Description of contents 23
insight and rigour if its problems are translated into the framework of
illocutionary semantics.
Chapter 8, 'Interjections across cultures', argues that interjections -
like any other linguistic elements - have meanings of their own, and
that these meanings can be identified and captured in the natural
semantic metalanguage. A number of interjections from English, Polish,
Russian, and Yiddish are discussed, and rigorous semantic formulae are
proposed which can explain both the similarities and the differences
in their range of use. For example, the English interjection yukI is
compared and contrasted with its nearest Polish and Russian counterparts
Jul, JeI, tful and t'Jul It is shown that while the meaning of interjections
cannot be adequately captured in terms of emotion words such as
disgust, it can be captured in terms of more fine-grained components,
closer to the level of universal semantic primitives. The role of sound
symbolism in the functioning of interjections is discussed, and the possi-
bility of reflecting this symbolism in semantic formulae is explored.
Chapter 9, 'Particles and illocutionary meanings', examines a number
of English and Polish particles, quantitative (for example only, merely,
just) and temporal (for example already, still, yet), and in each case
offers a paraphrase in natural semantic metalanguage, substitutable in
context for the particle itself. Special attention is given to 'approxima-
tive' particles, such as almost, around, about, or at least. It is shown
how the 'radically pragmatic' approach to the study of such particles,
advocated by Sadock (1981) and others, fails to account for the range of
their use. It is demonstrated that even the vaguest 'hedges' and 'approxi-
matives' (for example roughly and approximately) can be given rigorous,
and yet intuitively clear, semantic explications, which can explain their
uses, and the differences in the use of closely related particles, both
within a language and between different languages.
Chapter 10, 'Boys will be boys: even 'truisms' are culture-specific',
develops more fully a critique of a 'Gricean' or 'radically pragmatic'
approach to language use. Evidence against this approach is drawn
mainly from the area of colloquial 'tautologies' such as War is war or A
promise is a promise, which have often been adduced, by Grice and by
others, in support of such a 'radically pragmatic' approach to language
use. The chapter shows that such 'tautological constructions' are partly
conventional and language-specific, and that each such construction has
a specific meaning, which cannot be fully predicted in terms of any
universal pragmatic maxims. It is argued that the attitudinal meanings
conveyed by various tautological constructions and by similar linguistic
24 Introduction: semantics and pragmatics
devices can be stated in rigorous and yet self-explanatory semantic
formulae. 'Radical pragmatics' is rejected as a blind alley, and an
integrated approach to language structure and language use is proposed,
based on a coherent semantic theory, capable of representing 'objective'
and 'subjective' aspects of meaning in a unified framework.
Chapter 11, 'Conclusion: semantics as a key to cross-cultural pragmat-
ics', recapitulates the main features of the approach to the study of
human interaction advanced in the present book, stressing in particular
its universal, 'culture-free' perspective, and its 'multicultural', culture-
specific, content. It highlights the theoretical and methodological novelty
of the book, its empirical orientation, and its potential for use in
language teaching and in the teaching of cross-cultural understanding
and cross-cultural communication.
Chapter 2
Different cultures, different languages,
different speech acts
From the outset, studies in speech acts have suffered from an astonishing
ethnocentrism, and to a considerable degree they continue to do so.
Consider, for example, the following assertion: "When people make
requests, they tend to make them indirectly. They generally avoid im-
peratives like Tell me the time, which are direct requests, in preference
for questions like Can you tell me the time? or assertions like [' m trying
to find out what time it is, which are indirect requests." (Clark - Schunk
1980:111)
It is clear that these authors have based their observations on English
alone; they take it for granted that what seems to hold for the speakers
of English must hold for 'people generally'. Another author writes:
The focus of this chapter is on the situational conventions that influence
how people make, understand, and remember requests. I will argue that
people's knowledge of particular social situations results in certain re-
quests being seen as conventional. ... My starting point will be to show
how social contexts constrain the ways in which people comprehend
indirect requests.... I will sketch a new proposal that specifies how the
structure of social situations directly determines the surface forms used by
speakers in making requests. (Gibbs 1985:98)
This author seems to be quite unaware that there are people other than
speakers of English; consequently, he doesn't even suspect that 'surface
forms used by speakers in making requests' may differ from language to
language, and that if they do differ then they cannot be 'directly' deter-
mined by 'social situations'.
Throughout this chapter, I will try to show that statements such
as those quoted above are based on an ethnocentric illusion: it is not
people in general who behave in the ways described, it is the speakers
of English.
Presumably, the ethnocentric bias characteristic of speech act studies
is largely due to their origin in linguistic philosophy rather than in
linguistics proper (see below, section 5). Nonetheless, statements mistak-
26 Different cultures, different languages, different speech acts
ing Anglo-Saxon conversational conventions for 'human behaviour' in
general abound also in linguistic literature. I will quote just one more
characteristic example: "Every language makes available the same set of
strategies - semantic formulas - for performing a given speech act. ...
if one can request, for example, in one language by asking the hearer
about his ability to do the act (Can you do that?), by expressing one's
desire for the hearer to do the act (I'd really appreciate if you'd do that),
. .. then these same semantic formulas - strategies - are available to
the speakers of every other language." (Fraser - Rintell - Walters
1980:78-79). These authors are not unaware of some crosslinguistic
differences in this respect, but they dismiss them as 'minimal'.
Such preconceptions could probably be seriously dented by reference
to almost any language. Here, I shall be drawing mainly upon illustrative
material from Polish and from Australian English.
But even if one limits the task at hand to comparing selected speech
acts from only two languages, the topic is still vast and couldn't be
treated exhaustively in anyone work. The cultural norms reflected in
speech acts differ not only from one language to another, but also from
one regional and social variety to another. There are considerable
differences between Australian English and American English,
between mainstream American English and American Black English,
between middle-class English and working-class English, and so on.
There is also a great deal of variation within Polish. Nonetheless, there
is also a remarkable amount of uniformity within English, as there is
within Polish.
It goes without saying that the differences between English and Polish
discussed in this chapter could, and should, be studied in a much more
thorough and systematic way than has been done here. But to do so, one
would have to devote a whole book to the subject, or one would have to
limit one's field of vision to a strip so narrow that one would have no
grounds for reaching the generalisations which in my view explain
phenomena of the kind discussed here. The present overview was com-
piled as a pilot study. I believe, however, that even in its present form it
amply demonstrates that different cultures find expression in different
systems of speech acts, and that different speech acts become entrenched,
and, to some extent, codified in different languages.
Preliminary examples and discussion 27
1. Preliminary examples and dicussion
At a meeting of a Polish organisation in Australia a distinguished
Australian guest is introduced. Let us call her Mrs. Vanessa Smith. One
of the Polish hosts greets the visitor cordially and offers her a seat of
honour with these words:
Mrs. Vanessa! Please! Sit! Sit!
The word Mrs. is used here as a substitute for the Polish word pani,
which (unlike Mrs.) can very well be combined with first names. What is
more interesting about the phrasing of the offer is the use of the short
imperative Sit!, which makes the utterance sound like a command, and
in fact like a command addressed to a dog.
The phrase Sit down! would sound less inappropriate, but in the con-
text in question it would not be very felicitous either: it still would not
sound like an offer, let alone a cordial and deferential one. A very
informal offer could be phrased as Have a seat, with imperative mood,
but not with an action verb in imperative mood. More formal offers
would normally take an interrogative form:
Will you sit down?
Won't you sit down?
Would you like to sit down?
Sit down, won't you?
In fact, even very informal offers are often performed in English by
means of sentences in the interrogative form:
Sure you wouldn't like a beer? (Hibberd 1974:218)
Like a swig at the milk? (Hibberd 1974:213)
Significantly, English has developed some special grammatical
devices in which the interrogative form is normally used not for asking
but for making an offer, a suggestion or a proposal, especially the form
How about a NP?:
How about a beer? (Buzo 1979:64)
How about a bottle? (Hibberd 1974:187)
In Polish, How about utterances have to be rendered in a form indis-
tinguishable from that of genuine questions (except of course for the
intonation) :
28 Different cultures, different languages, different speech acts
Moze ezegos napijesz?
'Perhaps you will drink something?'
A further difference between Polish and English concerns the literal
content of interrogative offers. In English, a tentative offer (even a very
informal one) tends to refer to the addressee's desires and opinions:
Like a swig at the milk? (Hibberd 1974:213)
Sure you wouldn't like a bash at some? (Hibberd 1974:214)
The phrasing of such offers implies that the speaker is not trying to
impose his will on the addressee, but is merely trying to find out what
the addressee himself wants and thinks.
In Polish, literal equivalents of offers of this kind would sound inap-
propriate. The English question Are you sure?, so often addressed by
hosts to their guests, sounds comical to the Polish ear: it breaks the
unwritten law of Polish hospitality, according to which the host does not
try to establish the guest's wishes as far as eating and drinking is con-
cerned but tries to get the guest to eat and drink as much as possible (and
more). A hospitable Polish host will not take 'No' for an answer; he
assumes that the addressee can have some more, and that it would be
good for him or her to have some more, and therefore that his or her
resistance (which is likely to be due to politeness) should be disregarded.
A reference to the addressee's desire for food is as inappropriate in an
offer as a reference to his or her certainty. Sentences such as:
Miatbys oehotf na piwo?
'Would you like a beer?'
would be interpreted as questions rather than as offers. It would not be
good manners to reveal to the host that one feels like having a beer; the
social convention requires the host to prevail upon the guest, to behave
as if he or she was forcing the guest to eat and drink, regardless of the
guest's desires, and certainly regardless of the guest's expressed desires,
which would be simply dismissed. The typical dialogue would be:
Proszf bardzo! Jeszcze
Ale juz nie mogf(!
Ale koniecznie!
'Please! A little more!'
'But I can't!'
'But you must!' (literally: 'But necessarily! ')
Preliminary examples and discussion 29
What applies to offers applies also, to some extent, to invitations. For
example, in English a man can say to a woman:
Would you like to come to the pub tomorrow night with me and
Davo? (Buzo 1979:60)
Would you like to come out with me one night this week?
(Hibberd 1974:214)
Hey, you wouldn't like to come to dinner tonight, would you?
(Hibberd 1974:193)
In Polish, literal translations of such utterances would make very poor
invitations. A sentence in the frame:
Czy mialabys ochotf ?
'Would you like to ?'
sounds like a genuine question, not like an invitation or a proposal. If
a man wants to ask a woman out, it would sound presumptuous for him
to express overtly an assumption that she 'would like' to do it. Rather,
he should show that he would like to go out with her, and seek her
consent. One would say:
Moiebysmy poszli do kina?
'Perhaps we would go to the cinema?' (implied: if I asked you)
rather than:
Czy mialabys o o t ~ pojsc ze mng do kina?
'Would you like to go to the cinema with me?'
A tentative and self-effacing invitation such as the following one:
Say, uh, I don't suppose you'd like to come and have lunch with
me, would you? (Buzo 1974:44)
could not be translated literally into Polish without losing its intended
illocutionary force:
Powiedz, hm, nie przypuszczam, iebys miata ochote zjfsc lunch
ze mnf/, co?
The sentence sounds bizarre, but if it could be used at all it would be
used as a genuine question, not as an invitation or proposal. A question
of this kind could of course be interpreted as a prelude to an invitation,
but it would have to be reported as he asked me whether, not as he
invited me to. Clearly, one factor responsible for this difference is the
30 Different cultures, different languages, different speech acts
principle of 'polite pessimism', characteristic of Anglo-Saxon culture
(cf. Brown - Levinson 1978:134-135), but absent from Polish culture.
2. Interpretive hypothesis
Of course, Polish is not alone among European languages in differing
from English in the ways indicated above. On the contrary, it is English
which seems to differ from most other European languages along these
lines. Many of the observations made in the present chapter would also
apply to Russian, Serbo-Croatian, Spanish and many other languages. It
is English which seems to have developed a particularly rich syste.m of
devices reflecting a characteristically Anglo-Saxon cultural tradition:
a tradition which places special emphasis on the rights and on the auton-
omy of every individual, which abhors interference in other people's
affairs (It's none of my business), which is tolerant of individual idiosyn-
crasies and peculiarities, which respects everyone's privacy, which
approves of compromises and disapproves of dogmatism of any kind.
The heavy restrictions on the use of the imperative in English and the
wide range of use of interrogative forms in performing acts other than
questions, constitute striking linguistic reflexes of this socio-cultural
attitude. In English, the imperative is mostly used in commands and in
orders. Other kinds of directives (i.e., of speech acts through which the
speaker attempts to cause the addressee to do something), tend to avoid
the imperative or to combine it with an interrogative and/or conditional
form. (For certain important qualifications to this overall tendency, see
Lakoff 1972; Ervin-Tripp 1976.)
At least this is how English strikes native speakers of a language like
Polish, where the bare imperative is used on a much wider scale. It is
interesting to note that from a different cultural perspective English may
be seen as a language favouring, rather than shunning, the use of impera-
tive. This is, in particular, how English appears to speakers of Japanese.
For example, Higa (1972:53) notes the wide use of the imperative in
the English advertising language and points out that, for example, the
Japanese sign corresponding to the ubiquitous English Drink Coca-Cola!
would read Coca Cola 0 nomimasho! (Literally, 'We will drink Coca
Cola! ') rather than the imperative Coca Cola 0 nome! Similarly,
Matsumoto (1988:420) points out that in Japanese recipes or instructions
Case studies 31
an imperative would be avoided, whereas in English recipes or instruc-
tions it is quite common.
It should be noted, however, that advertisements and recipes are, first,
anonymous, and second, directed at an imaginary addressee, not at a
particular individual. What Anglo-Saxon culture abhors is the impression
that one individual is trying to impose his or her will upon another
individual. In the case of 'public speech acts' such as advertisements or
recipes this danger does not arise, and the imperative is not felt to be
offensive. In Polish, however, 'private' speech acts, directed from one
person to another, can also use the imperative, and they do not rely on
interrogative devices in this area either.
In what follows, I will consider a number of areas where Polish, and
other languages, differ from English along the lines suggested here,
specifically: advice, requests, tag questions, opinions, and exclamations.
3. Case studies
3.1. Advice
In a language like Polish, advice is typically offered in the form of
an imperative:
fa ci r d z ~ powiedz mu p r w d ~
'I advise you: tell him the truth.'
In English advice would normally be formulated more tentatively:
If I were you I would tell him the truth.
Tell him the truth - I would.
Why don't you tell him the truth? I think it would be best.
Why not tell him the truth? I think that might be best.
Maybe you ought to tell him the truth?
Do you think it might be a good idea to tell him the truth?
All these utterances could be reported in English using the verb advise
(She advised me to tell him the truth). But their literal Polish equivalents
would not be reported using the verb radzic 'advise'. Normally, only
utterances in the imperative mood or utterances with the verb radzit used
performatively could be so reported:
32 Different cultures, different languages, different speech acts
ci, zebys mu powiedzial
'I advise you to tell him the truth.'
It is also worth noting that the English verb advise is seldom used
performatively in ordinary speech: the phrase I advise you sounds very
stiff and formal; by contrast, its Polish equivalent ja ci sounds
perfectly colloquial and is frequently heard in everyday conversations.
3.2. Requests
In English, if the speaker wants to get the addressee to do something and
does not assume that he could force the addressee to do it, the speaker
would normally not use a bare imperative. Speech acts which could be
reported by means of the verbs request or ask (to) frequently have an
interrogative or an interrogative-cum-conditional form, as in the follow-
ing examples (all from Green 1975:107-.130):
Will you close the door please?
Will you close the window please.
Will you please take our aluminium cans to the Recycling Centre.
Would you take out the garbage please.
Would you get me a glass of water.
Would you mind closing the window.
Would you like to set the table now.
Won't you close the window please.
Do you want to set the table now?
Why don't you clean up that mess.
Do you want to get me a scotch.
Why don't you be nice to your brother for a change.
Why don't you be quiet.
Why don't you be a honey and start dinner now.
Not a single one of these utterances could be translated literally
into Polish and used as a request. In particular, literal equivalents of
sentences in the frame Why don't you would be interpreted as a combina-
tion of a question and a criticism, rather like utterances based on the
modal Why do it are in English (Why paint your house purple?) (See
Gordon - Lakoff 1975:96; cf. also Wierzbicka 1988:28.) In fact, a
sentence such as:
Case studies 33
Dlaczego nie zamkniesz okna?
(Literally) 'Why don't you close the window?'
would imply unreasonable and stubborn behaviour on the part of the
addressee ('why haven't you done what was obviously the right thing
to do - you should have done it long ago; I can't see any excuse for
your failure to have done it'). The corresponding English sentence could
also be interpreted in this way, but it doesn't have to be. In particular,
as pointed out to me by Jane Simpson (p.c.), the contracted from
Why' n' tcha suggests a request rather than a question.
It is worth noting in this connection that English has developed some
special devices for expressing requests and other directives in a partly
interrogative style, especially the expression Why don't you be (ADJ) ,
which can hardly be used for genuine questions. As pointed out in
Green (1975:127), the sentence Why aren't you quiet? can be a genuine
question, but the sentence Why don't you be quiet?! cannot. Thus, the
construction Why don't you be (ADJ)? has an interrogative form, and
an interrogative component in its meaning, but is specialised in speech
acts other than questions.
Characteristically, Polish has no similar constructions. Since in Polish
the use of interrogative forms outside the domain of questions is very
limited, and since the interrogative form is not culturally valued as a
means of performing directives, there was, so to speak, no cultural need
to develop special interrogative devices for performing speech acts other
than questions, and in particular, for performing directives.
As for literal equivalents of sentences in the frame Won't you, such as:
Nie zamkniesz okna?
'Won't you close the window?'
they would be interpreted as surprised questions (not necessarily critical
questions, but surprised questions). They would invite both an answer
and an explanation ('You are not going to do it? That's strange; I
wonder why?').
The difference between English and Polish in this respect becomes
particularly clear in cases of transference. For example, my daughters,
who are bilingual, but who live in an English-speaking environment,
often phrase their Polish requests interrogatively (or did when they
were younger):
Mamo, czy podasz mi
'Mum, will you give me a Kleenex?'
34 Different cultures, different languages, different speech acts
This sounds very odd to me, and I tend to correct them, urging them
to use the imperative (with the word r s z ~ 'please') instead. To an
English speaker, this might look like an attempt to teach one's child to
be impolite. But in Polish, politeness is not linked with an avoidance of
imperative, and with the use of interrogative devices, as it is in English.
The expression Would you mind has simply no equivalent in Polish. I
do not wish to imply, however, that Polish never uses the interrogative
form in requests. It does, but in comparison with English, the possibili-
ties are heavily restricted. Thus, one could perform requests, or acts
closely related to requests, by ostensibly 'asking' about the addressee's
ability to do something, or about his or her goodness (or kindness):
Czy mogfbys ?
'Could you ?'
Czy bylbys tak dobry, zeby ... ?
'Would you be so good as to ?'
Czy byf(a)by Pan(i) laskaw(a) ?
'Would you be so kind/gracious as to ... ?'
But one could not ask people to do something by using literal Polish
equivalents of the phrases Would you do it, Won't you do it, Why
don't you do it, Do you want to do it or Would you like to do it.
Pseudo-questions which ostensibly inquire about the addressee's desire
and which in fact are to be interpreted as requests (Would you like to, Do
you want to) seem particularly odd and amusing from a Polish point
of view, as transparent acts of what looks like naive hypocrisy.
But it is not just the range of acceptable interrogative devices which
distinguishes Polish directives from the English ones. Differences in
function are at least as striking. Thus, in Polish interrogative directives
sound formal and elaborately polite. They are also tentative, lacking in
confidence. One would use them when one is genuinely not sure whether
the addressee would do what is requested. Moreover, they could not be
used in anger (unless sarcastically) and they are incompatible with the
use of swear words. In Australian English, however, both the interroga-
tive and the interrogative-cum-conditional forms are frequently used in
speech acts which could be reported by means of the verbs order to,
command or tell to, and they are perfectly compatible with verbal abuse
and verbal violence, as the following examples demonstrate:
Case studies 35
Can't you shut up? (Hibberd 1974:228)
Why don't you shut your mouth? (Hibberd 1974:228)
Will someone put the fucking idiot out of his misery? (William-
son 1974:48)
Will you bloody well hurry up! (Williamson 1974:56)
For Christ's sake, will you get lost. (Williamson 1974:191)
Why don't you shut up? (Buzo 1979:37)
Andrew (to Irene, very angry): Will you please go to bed? (Wil-
liamson 1974:197)
Could you try and find the source of that smell before then, and
could you possibly put your apple cores and orange peel in the
bin for the next few days? (After a pause, loudly) And could you
bloody well shit in the hole for a change? (Williamson 1974:7)
In fact, the interrogative form in English has reached the stage of
being so thoroughly dissociated from the language of courtesy and re-
spect that it can well be used in pure swear phrases, where the speaker
forcefully expresses his feelings apparently without attempting to get the
addressee to do anything, as in the following example:
Why don't you all go to hell! (Hibberd 1974:199)
This shows particularly clearly that the English predilection for the inter-
rogative form in human interaction, and the heavy restrictions which
English places on the use of the imperative, cannot be explained simply
in terms of politeness. After all, Polish, too, has its polite and extra-
polite ways of speaking, and has developed a repertoire of politeness
devices. What is at issue is not politeness as such, but the interpretation
of what is socially acceptable in a given culture. For example, Australian
culture is highly tolerant of swearing. Swear words are often used to
express strong feelings and not only negative but also positive feelings,
as in the following examples:
Stork: Not bloody bad, is it?
Clyde: It's a bloody beauty. (Williamson 1974:18)
Bloody good music! (Buzo 1979:30)
36 Different cultures, different languages, different speech acts
There is no longer any widely shared taboo against swear words in
'polite conversation', for example in conversation with ladies about
music. On the other hand, there is evidently a strong reluctance to use
bare imperatives - not only in polite conversation, but even in not-so-
polite conversation. The implicit cultural assumption reflected in English
speech seems to be this: everyone has the right to their own feelings,
their own wishes, their own opinions. If I want to show my own feel-
ings, my own wishes, my own opinions, it is all right, but if I want to
influence somebody else's actions, I must acknowledge the fact that
they, too, may have their feelings, wishes or opinions, and that these
do not have to coincide with mine.
It is interesting to note that the flat imperative, which in English
cultural tradition can be felt to be more offensive than swearing, in
Polish constitutes one of the milder, softer options in issuing directives.
When the speaker gets really angry with the addressee, the speaker will
often avoid the imperative and resort to 'stronger' devices, in particular
the bare infinitive:
Nie pokazywac mi s i ~ tuta}!
'Not to show oneself to me here!' (i.e. 'You are not to come here.')
Wynosic s i ~ stqd!
'To get away from here!' (i.e. 'Get away from here! ')
Zabierac s i ~ stgd!
'To take oneself off from here!' (i.e. 'Off with you! ')
In the examples above (taken from Andrzej Wajda's film "Moralnosc
pani Dulskiej", based on a number of Gabriela Zapolska's plays), the
verbs chosen (wynosic s i ~ zabierac s i ~ are offensive and pejorative,
but especially offensive is the impersonal syntactic construction, with the
infinitive used instead of the more neutral imperative. The impersonal
infinitive seems to annihilate the addressee as a person (the absence of a
mention of the addressee in the sentence being an icon of his/her 'non-
existence'): it implies that the addressee is not worthy to be addressed
as an individual human being, and that the speaker does not wish to
establish any 'I-you' relationship with him/her. In particular, the speaker
excludes the possibility of any reply from the addressee. The infinitive
signals: 'No discussion' ('there is no person here whom I would regard
as a potential interlocutor, for example, as someone who could refuse or
decline to do as I say').
Case studies 37
By contrast, the English interrogative directives explicitly invite a
verbal response, as well as a non-verbal one (Okay, All right, Sure, and
the like), and thus indicate that the speaker views the addressee as an
autonomous person, with his or her own free will, who can always
decline to comply. The imperative is neutral in this respect: it neither
precludes nor invites a verbal response. Partly for this reason, no doubt,
it is favoured in Polish and disfavoured in English.
I would add that the infinitive construction is by no means restricted
to contexts where the speaker is angry. It can also be used simply to
assert one's authority; for example it can be used by parents who wish
to sound stern, as in the following example:
Macie parasol? Isc prosto - nie oglpdac
skromnosc - skarb dziewcz{!cia. (Zapolska 1978:30)
'Do you have the umbrella? (To) go straight - not to look
around. (To) remember: modesty is a girl's treasure.'
When the speaker wants to be more polite while still wishing to signal
coldness and a lack of intimacy, the infinitive can be used in combina-
tion with a performatively used verb:
do tego nie mieszac. (Zapolska 1978:108)
'I ask not to interfere.'
- powiedziec, nie krepowac. (from the
film "Moralnosc pani Dulskiej")
'I ask - I ask to say, I ask not to be embarrassed.'
In a sense, the infinitive directive functions as a distance-building device
in Polish, just as an interrogative directive does in English. But in
Anglo-Saxon culture, distance is a positive cultural value, associated
with respect for the autonomy of the individual. By contrast, in Polish
culture it is associated with hostility and alienation.
3.3. Tags
The deep-rooted habit of acknowledging possible differences between
individual points of view is particularly clearly reflected in the English
tag questions. Seen from a Polish point of view, English speech is
characterised by an all-pervasive presence of tag questions, highly
diversified in form and function. Essentially, Polish has only five or
38 Different cultures, different languages, different speech acts
six words which can be used as tags: prawda? 'true?', nie? 'no?',
tak? 'yes?', co? 'what?', dobrze? 'good', and nieprawdaz? 'not true?'
(slightly archaic). These are comparable to the English tags okay?,
right?, and eh? (this last one frequently encountered in Australia).
If these five or six Polish words were used nearly as often as English
tag questions are, Polish speech would sound grotesquely repetitive. The
English strategy of using auxiliary verbs - any auxiliary verbs, in any
combinations of moods, tenses and persons - as tags, ensures great
formal variety of tag questions. Expressions such as did he, was she,
have you, aren't they and so on may all have the same function, but the
sheer variety of their form allows them to be used much more frequently
than the five Polish tag words could be used.
But the differences between the English and the Polish systems
of tag questions go much further than that. The topic is vast and
obviously cannot be treated exhaustively here (see Chapter 6, section 5
on the illocutionary force of tag questions). Let me simply make a
few observations.
As has often been noted, English imperatives allow not one tag but
several, each with a slightly different function:
Close the door, will you?
Close the door, won't you?
Close the door, could you?
Close the door, can't you?
Close the door, why don't you?
Close the door, why can't you?
Close the door, would you?
In Polish, all these different tags would have to be rendered by means
of a single one: dobrze? 'well (good)?':
Zamknij drzwi, dobrze?
Semantically, the Polish tag corresponds most closely to the English will
you, the tag which assumes and expects compliance. The sentence Sit
down, will you? is more confident, more self-assured than Sit down,
won't you?, and the sentence Shut up, will you? sounds much more
natural than Shut up, won't you? Shut up, won't you could of course be
used sarcastically, but the sarcasm would exploit the effect of the
semantic and stylistic clash between the forcefulness of shut up and
the tentativeness of won't you.
Case studies 39
In contrast to won't you, will you can be used very widely, for
example in orders and commands, as well as in requests, and it is com-
patible with the use of swear words:
Look at this bloody ring, will you? (Williamson 1974:58)
So just move out, will you? (Buzo 1979:73) (said by a wife
throwing her husband out of their house)
In Polish in similar circumstances a bare imperative would normally be
used, unembellished by any tag whatsoever.
There are many other kinds of contexts where a tag question would
be used in English but not in Polish. In particular, English negative
questions with an opposite polarity would normally be translated into
Polish without a tag:
I don't suppose you've seen Hammo around, have you? (Buzo
1979:79)
Nie widziales przypadkiem Hammo?
(literally: 'You haven't seen Hammo by any chance?')
You are not having a go at me, are you? (Buzo 1979:11)
ezy ty ~ przypadkiem ze mnie nie nabijasz?
(literally: 'You are not having a go at me by any chance?')
You haven't heard anything about me, have you: Any sort of ...
rumours, have you? (Buzo 1979:64)
Nie slyszeliscie przypadkiem czegos 0 mnie? Jakichs ... plotek?
(literally: 'You haven't heard anything about me, by any chance?
Any rumours?')
Another situation where a tag question sounds plausible in English but
not in Polish can be illustrated with the following utterance:
I've made a bloody fool of myself, haven't /? (Williamson
1974:48)
The speaker discovers something about himself that he supposes the
addressees have been aware of all along. In Polish, a plausible thing to
say in a case like this would be widz{? 'I see', without a tag:
Widz{?, ie ~ zachowalem jak duren! (?co, ?prawda, ?tak, ?nie,
etc.)
'I see I have acted like a fool!' (?what, ?true, ?yes, ?no, etc.)
40 Different cultures, different languages, different speech acts
Again, I am not suggesting that tag questions are always used in
English out of consideration for other people or out of politeness. In
fact, they can be combined with accusations, insinuations and abuse,
as in the following examples:
Well. We have become a sour old stick, haven't we? (Williamson
1974:195)
What? You've changed your mind again, have you? (Williamson
1974:198)
You are a smart little prick, aren't you. (Williamson 1974:192)
You've engineered this whole deal, haven't you? (Williamson
1974:193)
You'd rather I was still over there, wouldn't you? (Williamson
1974:187)
In cases like these, one would not use a tag in Polish. In Polish the use
of tags is, by and large, restricted to situations when the speaker really
expects confirmation. In English, however, tag questions have come to
be so ubiquitous, and they have developed into such a complex and
elastic system, that their links with politeness, cooperation and social
harmony have become quite tenuous. Often, they are used as a tool of
confrontation, challenge, putdown, verbal violence and verbal abuse.
The very fact that tag questions have come to play such a major role in
English seems to reflect the same cultural attitudes which have led to the
expansion of interrogative forms elsewhere, and to the restrictions on
the use of the imperative, the same emphasis on possible differences of
opinion, of point of view. Basically, tag questions express an expectation
that the addressee will agree with the speaker, but the very need to voice
this expectation again and again signals constant awareness of a possibil-
ity of differences.
The range of contexts and situations where speakers of Polish would
invite confirmation is not nearly as wide, precisely because Polish
cultural tradition does not foster constant attention to other people's
'voices', other people's points of view, and tolerates forceful expression
of personal views and personal feelings without any consideration for
other people's views and feelings. In fact, the basic Polish tag, prawda?
'true?', presents the speaker's point of view not as a point of view
but as an objective 'truth'; and it doesn't seek agreement but an
acknowledgement of this 'truth'.
Case studies 41
Needless to say, it would be good if the observations ventured above
could be supported with text counts. So far, I have not undertaken any
large-scale counts of this kind. But to give the reader some idea of the
order of differences let me say, on the basis of a perusal of a large
anthology of Polish plays and of several volumes of Australian plays by
different authors, that one can easily get through fifty or more pages of
Polish plays without encountering a single tag, while in Australian
plays one can seldom get through five pages without encountering one,
and often one finds several on one page.
I would like to stress, however, that apart from quantitative differ-
ences suggested here, which require statistical validation, there are also
some indubitable qualitative differences. As a particularly clear example
I would mention chains of tag questions, characteristic of English
conversation but impossible in Polish. I quote a dialogue which I heard
not long ago at a bus stop in Canberra:
A: Lovely shoes, aren't they?
B: Aren't they nice?
A: Lovely, aren't they?
One might say that in exchanges of this kind the interlocutors are no
longer seeking confirmation, but rather are, so to speak, celebrating a
ritual of social harmony based on anti -dogmatism and religiously
respected freedom of judgement and right to one's own opinion.
Similarly, the difference between the 'opinion-oriented' English tag
('I think you would say the same; I don't know if you would say the
same') and the 'truth-oriented' Polish tag ('true?') is a matter of struc-
ture, not of frequency, and needs no statistical validation.
3.4. Opinions
In Polish, opinions are typically expressed fairly forcefully, and in every-
day speech they tend not to be distinguished formally from statements
of fact. One tends to say:
To dobrze. To niedobrze.
'That's good.' 'That's bad.'
as one says: 'That's white', 'That's black', in situations where in English
one would say: I like it, I don't like it, or even I think I like it.
42 Different cultures, different languages, different speech acts
As mentioned above, this difference is manifested in the structure of
Polish tag questions. One says in Polish, literally:
'She is nice (terrific), true?'
as if being nice or terrific or not were a matter of truth. In English, one
might say:
She is Italian, right?
but hardly
?She is nice, right?
??She is terrific, right?
But in Polish, the same tag, prawda 'true', would be used in both cases.
In Polish, one seldom presents one's opinions as just opinions (rather
than as 'the truth'), and one seldom prefaces them with expressions
such as I think, I believe or in my view. Expressions of this kind exist
of course (ja ja moim zdaniem, ja uwaiam) , but their use
is much more restricted than the use of their English equivalents. In
particular, Polish has no word which would correspond to the English
word reckon, which is used very widely in working class speech,
especially in Australia, in non-intellectual contexts, and which has no
intellectual pretentions. Translating utterances with I reckon into Polish
one would often have to leave it out, since all the conceivable Polish
equivalents would sound too intellectual, too cerebral, and simply would
not fit the context. For example:
Gibbo: I reckon it's the spaghetti they eat. Drives them round the
bend after a while. (Buzo 1974:37)
Jacko: (smiling) You know, Robbo, I reckon you'd have to be
about three hundred to have done all the things you reckon
you've done. (Buzo 1974:51)
Polish expressions such as Sf/,dze, or uwaiam would sound
as inappropriate in these contexts as the expressions I believe or in my
view would be in English. Similarly, the expression I guess, commonly
used in American English, is very colloquial, and it has no similarly
colloquial counterparts in Polish. In situations when in English one says,
for example:
I guess it's true.
Case studies 43
in Polish one would say simply:
To prawda.
'This is true.'
Drazdauskiene (1981) notes that expressions such as I think, I believe,
I suppose or I don't think are used much more often in English than they
are in Lithuanian. She suggests, basically correctly, I think, that they
signal "diminished assurance and therefore courteous detachment and
optional treatment of the subject matter" (1981 :57), and a desire not to
put one's view bluntly, and not to sound too abrupt or quarrelsome.
I don't agree, however, with her interpretation of this difference: "This
leads to a conclusion of the principal differential feature of English and
Lithuanian which is that in the familiar register English is verbally more
courteous and less straightforward than Lithuanian." (1981:60-61). In
my view, it is ethnocentric to say that Lithuanian is less courteous than
English (or, for a Lithuanian author, ethnocentric arebours): simply, the
rules of courtesy are different in each language. Furthermore, the signifi-
cance of the English norm in question should be seen as a reflection of
a deeper cultural attitude. English speakers tend to use expressions such
as I think or I reckon even in those situations in which they evidently
don't wish to be courteous, as in the following exchange:
Gibbo: Shows how much you know. Those back room boys work
harder than any of us.
Jacko: Ar bulls. I reckon it'd be a pretty soft cop being a back
room boy. (Buzo 1974:20)
As a different manifestation of the same cultural difference I would
mention the English preference for a hedged expression of opinions
and evaluations, and the Polish tendency to express opinions in strong
terms, and without any hedges whatsoever. Consider, for example, the
following exchange:
Norm: Well, you see, Ahmed, I'm all alone now, since my good
wife Beryl passed away to the heaven above.
Ahmed: I'm very sorry to hear that, Norm, you must feel rather
lonely. (Buzo 1979:15)
In Polish, one would not say anything like 'rather lonely'. Instead, one
would say bardzo samotny 'very lonely' or strasznie samotny 'terribly
lonely'. Similarly, if someone' s wife should kick him out of their
house, to live there with another man, it would be very odd to comment
44 Different cultures, different languages, different speech acts
on this situation in Polish using a term such as rather, as in the follow-
ing passage:
Richard: Tell me, how's your lovely wife?
Bentley: I don't know. She's living with Simmo in our home unit.
Richard: Bad luck.
Bentley: Yes, it is, rather. (Buzo 1979:64)
In English, hedged opinions go hand in hand with hedged, indirect
questions, suggestions or requests. People avoid making 'direct', force-
ful comments as they avoid asking 'direct', forceful questions or making
'direct', forceful requests. They hedge, and an expression such as rather
or sort of often fulfills a function similar to that of conditional and
interrogative devices. In fact, lexical hedges of this kind often co-occur
with grammatical devices such as the conditional and the interrogative
form, as in the following examples:
Richard: (to Sandy) Could you sort of ... put in a good word to
Simmo about me? (Buzo 1979:42)
Jacko: Oh, Pammy's a nice enough kid in her own way. But
you're sort of different. I mean, there's a lot more to you, I'd say.
I mean, now don't get me wrong, I'm not trying ... well, all I said
was, how about coming to lunch? (Buzo 1974:44)
Translating this last passage into Polish, one would have to leave out
several of the hedges. There is no way of saying I mean in Polish, in any
case no way of differentiating I mean from I' d say; there is no particle in
Polish which would correspond to well (cf. Wierzbicka 1976); and there
is no equivalent for sort of (except perhaps for jakas/jakos, but this is
closer to somehow than to sort of: the emphasis is on the speaker's
inability to describe the quality in question, not on a lack of full commit-
ment to what is said).
Thus, English is fond of understatement and of hedges; by contrast,
Polish tends to overstate (for emphasis) rather than understate. When I
translate my own writings from Polish into English, I find myself
removing words such as totally, utterly, extremely or always, or replac-
ing them with words and expressions such as rather, somewhat, tends to,
or frequently; and vice versa.
Case studies 45
3.5. Exclamations
The notion that English is fond of understatement is of course common-
place. Sometimes, however, the validity of this notion is disputed. For
example, it was questioned by Drazdauskiene (1981:66), who noticed
that strong positive stereotypical exclamations such as How lovely! or
Isn't it lovely! are much more common in English speech than they are in
Lithuanian speech. I would say that the same observation would apply to
Polish: Polish, like Lithuanian, makes frequent use of negative (critical)
exclamations but not of positive, enthusiastic ones.
I would point out, however, that the English understatement applies to
spontaneous opinions and feelings, not to opinions or feelings which are
presumed to be shared. The stereotypical exclamations discussed by
Drazdauskiene typically express enthusiastic appreciation for something
which the speaker presumes to be shared by the addressee. They often
sound exaggerated and insincere, and they certainly don't sound dog-
matic. The speaker is not bluntly stating hislher own view, disregarding
any potential dissent; on the contrary, he (or, according to the stereotype,
she) is eager to agree with the addressee. It is of course highly significant
that, as mentioned earlier, the stereotypical exclamations often take an
interrogative form (Isn't that lovely?) or are followed by a symmetrical
question asking for confirmation (How wonderful! Isn't that wonderful?)
Drazdauskiene suggests that the difference between English and
Lithuanian with respect to the use of stereotyped positive exclamations
may be related to the fact that Lithuanians are reserved and restrained
(and this view, expressed by a Lithuanian, certainly agrees with the
Polish stereotype of Lithuanians). But Poles, unlike Lithuanians, are not
regarded as restrained or reserved, and yet in this particular respect they
seem to be closer to Lithuanians than to speakers of English. I suggest
that exclamations under discussion do not point to any lack of emotional
restraint on the part of the speakers of English. On the contrary: they
are a conventional device aimed at 'being nice' to the addressee rather
than any spontaneous and unrestrained outburst of the heart.
In English, exclamations can take not only an affirmative and positive
form, as in:
How nice!
but also (especially in what tends to be regarded as more typically femi-
nine speech) an interrogative-negative one, as in the utterance:
46 Different cultures, different languages, different speech acts
Isn't he marvellous! (Buzo 1979:41)
Thus, the function of such exclamations is similar to that of tag questions
with an opposite polarity:
Terrible place, isn't it?
Negative-interrogative exclamations do not always have an interroga-
tive intonation, and do not always invite confirmation. Often, they are
used simply to express the speaker's feeling, and are followed by a
positive statement from the speaker rather than by a pause to be filled
by the addressee:
Bentley: Isn't she a sweetie? a real darling. (Buzo 1979:45)
Sundra: Wasn't that funny? That was the funniest thing I've ever
heard. (Buzo 1974:114)
Sundra: Isn't that nice of them? I think that's very nice of them.
(Buzo 1974:115)
Sundra: Isn't that wonderful? I think that's wonderful. (Buzo
1974: 115)
However, even when interrogative-negative exclamations are not used as
a truly dialogic device they still signal (at least in a perfunctory way) an
interest in what the addressee would say; they acknowledge the possibil-
ity that the addressee could say the opposite (even though the speaker
regards this as unlikely) and symbolically seek confirmation. The
speaker expects agreement, but does not take this agreement for
granted, and 'graciously' leaves the addressees the opportunity to
express their point of view, too. All this may of course be purely
perfunctory, purely conventional, but the convention is there, and it has
its own cultural significance.
Characteristically, in Polish there is no similar convention. Exclama-
tions always take a positive form:
Jak gJupo!
'How stupid!'
Wspaniale!
'Wonderful! '
The interrogative form would be interpreted as a genuine question.
Cultural values reflected in speech acts 47
4. Cultural values reflected in speech acts
4.1. Lexical evidence
The cultural differences between English and Polish discussed here have
also innumerable lexical reflexes. I will mention two of them here.
One is the presence in the English lexicon of the word privacy, which
has no equivalent in Polish, nor, apparently, in other European languages.
In fact, the concept of privacy seems to be a characteristically Anglo-
Saxon one. The word privacy is a very common one, frequently used in
everyday speech, and it clearly reflects one of the central values of
Anglo-Saxon culture. To have privacy means, roughly, 'to be able to do
certain things unobserved by other people, as everyone would want to
and need to'. The cultural assumption embodied in this concept is very
characteristic: it is assumed that every individual would want, so to
speak, to have a little wall around him/her, at least part of the time,
and that this is perfectly natural, and very important.
One is tempted to speculate, in this connection, that the absence of
an intimate T-form of address (in the sense of Brown - Gilman 1972),
which sets English apart from other European languages, is a reflex of
the same attitude. The English you is of course very democratic, it is a
great social equaliser, but it can also be seen as a distance-building
device. This is not to say that the meaning of the English word you is
analogous to that of a V-form in a language which does have a T-V
contrast. But I think that in the absence of such a contrast the form you
can't convey the intimacy signalled by the choice of a T-form. An
intimate form allows the speaker to get psychologically close to the
addressee, to penetrate the wall surrounding each individual. The English
you keeps everybody at a distance. In Anglo-Saxon culture non-sexual
body contact is heavily restricted, as compared, for example, with Slavic
and Mediterranean cultures: people seldom touch one another, hug one
another, kiss one another, or seldom even shake hands (see Triandis -
Triandis 1960). They also physically keep at a considerable distance
from one another, as compared, for example, with Slavs (cf. for example
Monahan 1983). The absence of an intimate T-form reflects and fosters
the culturally expected psychological distance between individuals, the
general need for psychological and physical 'privacy'.
One might add that the cultural taboo on 'personal remarks' character-
istic of Anglo-Saxon culture, and the existence in English of the set
48 Different cultures, different languages, different speech acts
expression personal remarks, with its negative implications, can be seen
as another strategy for building a little protective wall around every
individual. There is no similar expression in Polish, and there is no
similar taboo in Polish culture.
Of course English doesn't have the elaborate distance-building defer-
ential devices of Far Eastern languages such as Japanese, Korean,
Javanese or Thai, either. It is interesting to note that from the perspective
of languages of this kind English may appear as a language highly
sensitive to intimacy. (Cf. for example Hijirida - Sohn 1986:391.)
But this is an illusion. American (and generally English) address forms
such as Bob, Jim, Tom or Kate, have nothing to do with intimacy. It is
true that they imply less 'distance' than, for example, Dr Smith, but they
don't come anywhere near the potential intimacy of the Polish ty, or the
French tu, or even the Japanese (2SG) kimi. They imply informality and
friendliness, not intimacy. Intimacy implies an especially close personal
relationship between the speaker and the addressee; and English has no
devices to convey that. For example, at an Australian university a head
of department, or a dean, may send a memo to all the members of the
department or the faculty, signing it with a first-name form: Bob, or Bob
Johnson; and at a meeting of a university committee, the members from
different departments may introduce themselves to each other as Bob
(Johnson) and Kate (Brown), and start addressing each other as Bob and
Kate, from the very first meeting. This has nothing to do with intimacy.
They can be friendly, informal, and familiar, but they are not claiming,
in this way, any 'special relationship' with the addressee. (For further
discussion, see Chapter 3, section 3.)
The universal English you is of course less 'distant' than the deferen-
tial Japanese third person forms of address such as sensei 'teacher', or
than the deferential Polish third person form of address such as Pan
Profesor 'Mr Professor'; but it is also far less 'intimate' than the
Japanese kimi or the Polish ty. Being the great equaliser, the English you
keeps everybody at a distance - not a great distance, but a distance; and
it doesn't allow anybody to come really close.
The second lexical difference between English and Polish that I would
like to comment on concerns the concept embodied in the English word
compromise (in the sense of mutual concessions) and its Polish
counterpart, kompromis. In English, the word is essentially neutral, and
if it has any value connotations they would tend to be positive rather
than negative. By contrast, the Polish word tends to be used with
negative connotations. In any case, lexical and phraseological derivates
Cultural values reflected in speech acts 49
of kompromis unquestionably embody value judgments. Thus, pojse
na kompromis 'accept a compromise' suggests a moral weakness, a
deplorable lack of firmness, a sell-out of values. The adjective
bezkompromisowy 'without compromise' (said of someone who would
never accept a compromise) is emphatically positive: it is a word of
high praise, like heroic, noble or immaculate.
Thus, in the Polish cultural tradition, holding firmly to one's beliefs
and making no concessions to those of others is a valued and desirable
attitude. In the Anglo-Saxon tradition, similar attitudes would be
regarded as dogmatic and inflexible, and would be viewed with disap-
proval. In fact, the word inflexible and its Polish literal counterpart
n i u i ~ t y provide another example of the same kind: the English word
has negative connotations, whereas the Polish one is highly positive.
Polish has also the word niezJomny 'unbreakable', which is also a term
of praise and has no equivalent in English. (For further discussion see
Wierzbicka, to appear, chap. 6.)
4.2. Objectivism as a cultural value
The complex of cultural attitudes which conditions every individual to
be constantly aware of other people, other voices, other points of view,
to see oneself as one individual among many, all of them equally entitled
to their psychological space, their autonomy, their own peculiarities and
eccentricities, leads to objectivism and anti-dogmaticism being regarded
as important social and cultural values.
I would venture to suggest that this objectivism may be reflected in
peculiarly English ways of referring to oneself, and to one's own coun-
try, as it were from an external point of view. This can be illustrated with
the characteristic expression this country (commented on in Doroszewski
1938:120). In Polish, it would be inconceivable to refer to one's own
native land, one's ojczyzna 'fatherland' as ten kraj 'this country', as if
it were one among many countries, where one just happens to be at a
particular time. The Polish expression ten kraj could only be used with
respect to a foreign country; if it was used with reference to one's own
country it would mark the speaker as a psychological emigre.
Similarly, in English it is possible to refer to one's own nation as this
nation, especially in an elevated rhetorical style. In Polish one would say
in similar circumstances nasz nar6d 'our nation' (cf. Nasz narod jak
lawa, z wierzchu zimna i martwa, sucha i plugawa, 'Our nation is like
50 Different cultures, different languages, different speech acts
lava, at the surface cold and lifeless, dry and repellent,' Mickiewicz
1955:210). To say ten nar6d 'this nation', would indicate a complete
lack of identification with one's nation; to use this expression one would
have to psychologically leave one's own nation. As a further example,
consider the English expression (the) same here, referring to oneself,
as in the following dialogue:
Michael: I might just have a small claret.
Carmel: Same here. (Williamson 1974:155-156)
In Polish, the literal translation of same here or the same here would be
simply incomprehensible (as a way of identifying the speaker).
It seems to me that this inclination to look at oneself from outside, to
be conscious of the existence of many different points of view, all of
them equally valid (at least potentially), fits in very well with the other
characteristic features of English speech described here.
4.3. Cordiality as a cultural value
Throughout this chapter, the emphasis has been mainly on Anglo-Saxon
cultural values, reflected in the English language. Polish has been pre-
sented mostly in negative terms, as lacking certain devices characteristic
of English. I would like to say a few words to redress the balance. It
would be ridiculous to suggest that English speech acts reflect certain
cultural values whereas Polish speech. acts reflect nothing but an absence
of those values. It goes without saying that in fact Polish reflects
values characteristic of Polish culture. From an English speaker's point
of view, Polish ways of speaking may appear to reflect dogmatism, lack
of consideration for other people, inflexibility, a tendency to be bossy,
a tendency to interfere, and so on. On the other hand, from a Polish
speaker's point of view, English ways of speaking may be seen as
reflecting a lack of warmth, a lack of spontaneity, a lack of sincerity.
The central place of warmth, of affection, in Slavic as well as in
Mediterranean cultures, is reflected, among other things, in the rich
systems of expressive derivation, and in particular in the highly
developed systems of diminutives, involving not only nouns, but also
adjectives and adverbs. By contrast, in English, productive diminutive
derivation hardly exists at all, despite the existence of isolated baby
forms such as handies, doggie or birdie (one can say girlie but not
*mannie; auntie but not *unclie, horsie but not *goatie, and so on).
Cultural values reflected in speech acts 51
The central role of 'warmth', of affection, in Polish culture (and in
Slavic culture in general) is evidenced above all in the expressive
derivation of personal names (which goes much further than anything
one can find, for example, in Italian or Spanish). The topic is vast, and
cannot be discussed here in detail. Let me just mention that one personal
name, for example Anna or Maria, can have in Polish as many as
ten different derivates, all commonly used with respect to the same
person, each of them implying a slightly different emotional attitude,
and 'emotional mood'. For example:
Anna: Ania, Anka, Aneczka, Anusia, Anuska, Anusienka, Anulka,
Anuchna; Anusifltko
Maria: Marysia, Marysienka, Maryska, Marysiuchna, Marychna,
Marys, Marysiulka, Marycha, Marysifltko
This is quite apart from a variety of forms such as Maryla, Mania,
Marynia, Maryna, etc. (all from Maria), which are usually chosen from,
for a particular person, on a more permanent basis. (For a detailed
discussion of the semantics of expressive forms of names in Polish and
in Russian, see Wierzbicka, to appear, chap. 7.)
I would suggest that there are many subtle ways in which expressive
derivation interacts with speech acts. The topic deserves a separate study.
In this chapter, I will mention just two examples of this interaction.
In Polish, warm hospitality is expressed as much by the use of diminu-
tives as it is by the 'hectoring' style of offers and suggestions. Character-
istically, the food items offered to the guest are often referred to by the
host by their diminutive names. Thus, instead of asking:
Would you like some more herring? Are you sure?
one might say in Polish:
Wei jeszcze sledzika! Koniecznie!
'Take some more dear-little-herring-(DIM)! You must!'
The diminutive praises the quality of the food and minimises the quantity
pushed onto the guest's plate. The speaker insinuates: 'don't resist! it is a
small thing I'm asking you to do - and a good thing!' The target of the
praise is in fact vague: the praise seems to embrace the food, the guest,
and the action of the guest desired by the host. The diminutive and the
imperative work hand in hand in the cordial, solicitous attempt to get the
guest to eat more. Certainly, the cultural style of such offers is very
different from that of Would you like some more?, but the difference
52 Different cultures, different languages, different speech acts
cannot be described in terms of politeness. Rather, it has to be described
in terms of different cultural traditions, and, ultimately, different
hierarchies of values.
If one's own view of what is good for another person does not
coincide with the view of that person, Anglo-Saxon culture requires that
one should rather respect the other person's wishes (i.e., autonomy)
than to do what we think is good for the person; Polish culture tends
to resolve the dilemma in the opposite way.
A similar dilemma is involved in leave-taking behaviour: if the guests
indicate that they are about to leave, should one let them go or should
one try to prevent them from leaving? In Anglo-Saxon culture, one
usually lets them go, acknowledging in this way their autonomy and
'self-determination'. In Polish culture, however, such behaviour would
be seen as cold and uncaring; usually, therefore, one tries to prevent the
guests from leaving, since a display of warmth towards the addressees is
perceived as more important than a display of respect for their autonomy.
An Anglo-American or Australian host, therefore, would normally thank
the guests for coming and let them go, whereas a Polish host would insist
that the guests must stay longer, and would shower them with 'you
must's and with warm diminutives at the same time:
Ale jeszcze Ale koniecznie!
'But [stay] a little-DIM more! But you must!
As a third example of the interaction between diminutives and illocu-
tionary strategies I will mention requests. In Polish, a request formulated
in the imperative mood would often be softened by means of a diminu-
tive. Thus, while it would be more natural for a wife to use an imperative
than an interrogative-cum-conditional request when speaking to her
husband, she would be likely to soften that imperative by a double
diminutive form of his name (as well as by intonation):
Jureczku, daj mi papierosa!
'George-DIM-DIM, give me a cigarette!'
An indirect interrogative request would be less appropriate in this
situation because 'interrogativity' in directives is a distance-building
device: there is an implicit conflict between intimacy and affection on
the one hand and complete mutual independence on the other. (If I ask
you to do something for me, and if I think we are close, I will assume
that you will do what I want you to do; to show that I don't know if
Cultural values reflected in speech acts 53
you'll do it is to acknowledge your independence, but also, your
'distance' from me.)
Similarly, in speaking to a child one would be unlikely to use an
interrogative request (could you, would you be so good as to). Normally,
one would use an imperative. But this imperative would be likely to be
softened not only with a multiply diminutive form of the name, but also
with numerous other diminutives, on nouns, adjectives, adverbs, and
occasionally some other parts of speech:
Monisienko, jedz
'Monica-DIM-DIM, eat your soup-DIM!'
Jedz
'Eat quickly-DIM!'
Zjedz wszysciutko!
'Eat it all-DIM up!'
Rich systems of diminutives seem to playa crucial role in cultures in
which emotions in general and affection in particular is expected to be
shown overtly. Anglo-Saxon culture does not encourage unrestrained
display of emotions. In adult English speech diminutives (even those
few diminutives which English does have) feel out of place, just as
non-erotic kissing and hugging feels more often than not out of place.
It is fascinating to note, in this connection, that in comparison with
say Japanese culture, Anglo-Saxon culture in general and American
culture in particular emerges as one which greatly encourages physical
expressiveness. Barnlund (1975a:445) reports a "dramatic contrast
between the [American and Japanese] cultures" in this respect. "Touch-
ing behavior is reported nearly twice as often in all categories and with
all persons by Americans as by Japanese." (1975a:452). On the other
hand, American students of Russia and things Russian are amazed by the
amount of touching, kissing and hugging which visibly takes place
among the Russians (cf. Smith 1976:136; Monahan 1983).
From a Polish perspective, Anglo-Saxon culture in general (including
American culture) seems as restrained in physical expressiveness as
Japanese culture seems to Americans. Most observers seem to agree that
the Poles are not quite as effusive as the Russians, but, for example,
kissing, hand-kissing and hand-shaking in greetings take place on a
daily basis.
The overtones which the word emotional has acquired in English are
a good illustration of the disapproval of public display of emotions,
54 Different cultures, different languages, different speech acts
characteristic of Anglo-Saxon culture. Frequently this word is used with
negative connotations, but even when it is not it implies at least 'an
unexpected and somewhat embarrassing display of emotions' .
For example, when an abducted baby was returned, after two days, to
his mother, who thought she would never see him again, the Australian
reporter (ABC News, 24.8.1983) described the mother's behaviour as
'emotional' ("The baby was reunited with his emotional mother"). In this
particular context, the word emotional is not used as a criticism, since
the mother's 'emotional state' is apparently seen as something that can
be understood and excused. Nonetheless, from a Polish speaker's point
of view, the very need to mention and to excuse the mother's emotion
would seem odd: it would not occur to one that a mother could do other
than show emotion in a situation of that kind.
As pointed out by Lutz (1986:290-301) the "widely shared American
[and, I would add, Anglo-Saxon in general - A.W.] ethnotheory of
basically Protestant European, middle class background" identifies
"emotion primarily with irrationality, subjectivity, the chaotic and other
negative characteristics". "One of the most pervasive cultural assump-
tions about the emotional is that it is antithetical to reason or rationality";
"emotions are fundamentally devalued ... as irrational, physical, unin-
tentional, weak, biased, and female"; "emotions tend predominately [sic]
to lead to erroneous judgements and hence senseless or irrational actions.
. .. people tend to see emotion as a disruption of, or barrier to, the
rational understanding of events. To label someone emotional is often to
question the validity, and more, the very sense of what they are saying."
Not so in Polish culture. In the romantic poetry which played a funda-
mental role in shaping Polish national ethos, serce 'heart' is opposed
to the scientist's szkielko i oko 'magnifying glass and eye', as a source of
'live truth' versus the domain of 'dead truths', and this opposition has
retained an important place in the Polish ethnotheory. The fact that the
Polish counterpart of the English word emotional, that is, uczuciowy,
has positive connotations, reflects this. Uczuciowy does not designate
someone who shows emotion (because there is no cultural expectation
that feelings would or should not be shown), but rather someone who
possesses rich and strong emotions (seen as a 'good thing').
It must be stressed, however, that the Anglo-Saxon taboo on 'emo-
tions' does not concern all feelings to the same degree. For example, as
mentioned earlier, in Australian culture it is quite all right to swear, that
is to show 'strong', 'masculine' feelings. What is not all right is to show,
without restraint, 'weak', 'soft', 'feminine' emotions, such as tenderness.
Cultural values reflected in speech acts 55
Lutz (1986:299) points to the Anglo-Saxon (she says, American)
distinction between emotions seen as typical of, and forgivable in,
women, and those which can be expected of men. "American cultural
belief does not deny that men may become emotional; it does, however,
engender expectations that men will experience only certain types of
emotion, notably anger. Women are expected to experience the entire
range of emotions more frequently and deeply, with the possible excep-
tion of anger".
In Australian culture, which highly values 'toughness' and anti-
sentimentality, and where the word bloody is the main vehicle for
expressing emotions (both negative and positive ones), any display of
'soft', 'feminine' emotions is particularly abhorred. (Cf. Wierzbicka, to
appear, chap. 11.)
It is worth noting in this connection that characteristically Australian
abbreviations, such as mozzies (mosquitoes), mushies (mushrooms),
prezzies (presents), barbie (barbecue), lippie (lipstick), or sunnies (sun-
glasses), which are often referred to as diminutives, in fact are not really
diminutives and have a function quite different from the main function
of diminutives (although it is of course a simplification to speak of
diminutives as if they had only one function). Formally, they differ from
English diminutives because they are abbreviations: baby words such
as birdie, fishie or doggie add a diminutive suffix to the full form of the
base word, but words such as barbie or lippie add a suffix to a truncated
form of the base word. Semantically, they differ from diminutives in
expressing, essentially, not endearment but good humour. The core
meaning of true diminutives (such as doggie) can perhaps be represented
as follows (cf. Wierzbicka (1980, 1984; to appear):
doggie =>1
I think: this is something small like you are someone small
I feel something good towards you
because of this, when I say something about this to you
I feel something good towards it
The core meaning of Australian abbreviations with the suffix -ie is dif-
ferent. I would represent it as follows (cf. Wierzbicka 1984; to appear):
mozzies =>
I think: this is something small
I think: you think the same
when I say something about this to you I feel something good
56 Different cultures, different languages, different speech acts
Thus, calling mosquitoes mozzies, the speaker is good-humouredly
dismissing the problem; he thinks of mozzies as small-(but not endear-
ing), and expects that the addressee would share this attitude. As I have
suggested elsewhere, the semantic complex explicated above reflects
many characteristic features of the Australian ethos: anti-sentimentality,
jocular cynicism, a tendency to knock things down to size, 'mateship',
good-natured humour, love of informality and dislike for 'long words'
(Slavic or Romance diminutives are typically much longer than the
base words, but Australian abbreviations are normally shorter than the
base words, and Australians feel that this formal brevity is somehow
functional) .
As another linguistic reflex of the same Australian attitudes, and in
particular of the Australian non-sentimental good humour, I would
mention the quintessentially Australian expression no worries, which
permeates Australian speech and which serves a wide range of illocu-
tionary forces. The casual optimism encapsulated in this expression and
also in the Australian abbreviations is something quite different from
the warmth of Slavic diminutives.
4.4. Courtesy as a cultural value
I think it is important to add that while Polish culture shares one major
theme of Slavic culture in general, cordiality, it combines it with a differ-
ent one: courtesy, in the sense of a somewhat ceremonial show of respect
for every individual person (and especially for women). There is in
Polish culture, alongside cordiality and spontaneity, an element of
ceremony, of somewhat ritualised courtesy and chivalry. The Polish
custom of kissing a lady's hand (by men) is a characteristic example of
this: vigorous warm kisses on both cheeks signal cordiality, but one
kiss on a lady's hand signals both cordiality and ceremonial courtesy.
Courtesy is not in conflict with cordiality, but it imposes on it certain
ritual forms, a certain ceremoniality.
The courtesy aspect of the Polish savoir vivre is manifested particu-
larly clearly in forms of address.
2
As mentioned earlier, in English every-
body (except perhaps the Queen) can be addressed in the same way,
as you. In Polish, one always distinguishes the intimate ty 'thou' from
the courteous pan/pani 'sir', 'madam' (with the verb in third person
singular). The English you is democratic, the same for everyone; it lacks
Cultural values reflected in speech acts 57
both the (potential) intimacy of the Polish form ty and the courteousness
of the Polish forms pan/pani.
This link between courtesy and cordiality is interesting to note be-
cause it seems to be, typologically, rather unusual. Ceremony and ritual
may seem to be antithetical to spontaneity and 'emotionality', and cul-
tures which favour the former usually restrict the latter. The Japanese
and Javanese cultures are cases in point (see Benedict 1947; Lebra 1976;
Smith 1983 on Japanese culture; Geertz 1976 on Javanese).
But Polish culture distinguishes sharply between spontaneity and
emotionality on the one hand and informality on the other. Like
Japanese, Polish is very fond of titles, and the list of titles commonly
used goes far beyond the 'Doctor', 'Professor' or 'Father', commonly
used in English. For example, one says commonly Panie Dyrektorze
'Mr. Director', Panie Naczelniku 'Mr. Head', Panie Iniynierze 'Mr.
Engineer', Panie Magistrze 'Mr. MA-holder' (usually said to a pharma-
cist, who holds an MA in pharmacy), Panie Mecenasie 'Mr. Barrister',
and so on. But unlike in Japanese, in Polish the 'language of respect'
doesn't involve humility and self-abasement: one pays respect to the
status and rank of the addressee without ever lowering oneself. Further-
more, this respect for the addressee is commonly combined in Polish
with cordiality and affection.
The compatibility between courtesy and cordiality is best seen in
forms of address or of personal reference which combine formal titles
pan 'Mr.', pani 'Mrs.' and panna 'Miss' with affectionate diminutive
forms of personal names, such as Panie Mareczku 'Mr Mark-DIM' or
Pani Basienko 'Mrs Barbie-DIM'. Polish dislikes informality (which is
so characteristic, for example, of Australian English), and it encourages
the use of titles even between 'equals' who know each other very well,
and who have known each other for years (for example, between work-
mates). At the same time, however, the formality of such forms of
address does not prevent the show of emotion, and affectionate diminu-
tives of first names are freely combined with titles, as they are with
hand-kissing.
Polish differs in this respect from Russian, which has also a wealth of
devices for showing emotion, but which is not similarly rich in devices
for showing courtesy, and which links affection with informality. To
show respect, courtesy, and non-intimacy one uses in Russian a combi-
nation of full first name and patronymic, and normally the patronymic
cannot be combined with an affectionate diminutive. (Cf. Wierzbicka, to
appear, chaps. 7,8.)
58 Different cultures, different languages, different speech acts
Apart from names and names with patronymics, Russian has two basic
forms of address: ty 'thou' and vy 'you' (PL), one of which signals
'intimacy' and the other 'distance'. But the Russian form vy doesn't
correspond exactly to the Polish forms pan/pani, because it signals only
'distance', not courtesy. In Warsaw shops one sometimes encounters
Russian women tourists, delighted and amused to be addressed as
pani 'Madam' - a form which they perceive as quaintly courteous and
ancien regime.
The absence of a special courtesy value in the Russian form vy makes
it suitable for use among party apparat and police as well as among
ordinary people. In communist Poland, police and party apparat avoided
the forms pan/pani, whose 'aura' didn't fit the communist ideology.
Characteristically, the communist regime in Poland attempted for many
years to eradicate these forms, replacing them with wy (on the Russian
model). (Cf. Davies 1981,2:581.) These efforts, however, proved futile.
The fact that in the documentary film 'Workers 1980' the representatives
of the Government, talking to Lech l ~ s and other representatives of
the workers, used publicly the forms pan/pani, was widely commented
on in Poland, as a kind of symbolic recognition of the defeat of efforts
aiming at eradicating the Polish tradition of courtesy.
Following on Brown - Gilman (1972), different forms of address
such as ty vs. pan/pani in Polish are usually described in terms of 'power
and solidarity' (see, however, Ervin-Tripp 1974). I would suggest, how-
ever, that as far as Polish is concerned, it is more illuminating to refer
here to cultural values such as intimacy and courtesy. The forms pan/
pani differ from the so-called V-forms of languages such as Russian in
having positive courtesy built into them. The form wy (second person
plural), favoured by the communist regime, carried with it implications
of impersonal equality, as well as distance. To the Polish ear, it sounded
cold, impersonal and discourteous. It de-emphasised personal ties (either
intimate, signalled by ty, or based on mutual respect, signalled by pan/
pani) in favour of equality derived from membership in a collectivity.
Pan/pani, on the other hand, is non-intimate, but it is also courteous and
personal. I presume that the 'personal' character of pan/pani is due partly
to its singular form, and possibly also to its sex differentiation, whereas
the 'impersonal' character of the form wy is due partly to its plural and
genderless form. Polish courtesy stresses respect for every individual
as an individual, and is highly sex-conscious. The collectivist and
genderless ring of the form wy is jarring, in that tradition.
Theoretical implications 59
One might add that, in communist Poland, the officially-supported
form wy co-occurred with 'collectivist' vocatives and appellatives such
as towarzyszu, towarzysz 'comrade', and, to a lesser extent, obywatelu,
obywatel 'citizen', and was no doubt interpreted in conjunction with,
or against the background of, Soviet-style forms. The appeal to Soviet-
style equality conveyed by the official wy was backed by an explicit or
implicit reference to a collectivity of 'comrades', that is, ideologically
committed equals.
In Polish dialects, the form wy has a different origin and a different
function: it is opposed to ty only (not to ty on the one hand and to pan/
pani on the other), and it expresses not equality but respect. Signifi-
cantly, it doesn't co-occur there with any collectivist and ideologically
loaded forms of address such as towarzysz 'comrade'. Rather, it co-
occurs with terms referring to the addressee's personal status, such as
'mother' or 'uncle', or with first names, usually in a 'dignified', non-
diminutive form.
I would add that the contrast between the courteous, Polish-style form
pan/pani and the impersonal, Soviet-style form wy is something that
Poles are acutely aware of and often comment on. To illustrate this
general awareness of the semantic implications of the two forms, I quote
a characteristic passage from an essay which was published in the lead-
ing Polish emigre monthly, Kultura:
When the Russians speak of us ironically as te polskie pany ['those Polish
gentlemen'], the connotations are of culture rather than class. The gentry as
a class has long since ceased to exist, but we are still 'gentry' because we
didn't submit to Soviet attempts at 'Gleichschaltung', at 'comradising' us,
and the form wy ['you PL'] didn't take. In communist Poland the only
contrast really felt is that between panowie ['gentry', but also 'misters']
and those who are generally referred to as oni ['they', i.e. the regime
people, the new ruling class]. (Schrett 1984:7)
5. Theoretical implications
In the literature on speech acts, English conversational strategies dis-
cussed here are frequently interpreted as a manifestation of a universal
'natural logic' (Gordon - Lakoff 1975), a universal 'logic of conversa-
tion' (Grice 1975) or universal rules of politeness (Searle 1975). In the
60 Different cultures, different languages, different speech acts
light of the facts discussed in this chapter this line of interpretation
must be seen as highly ethnocentric. Consider for example, the following
statement:
. .. ordinary conversational requirements of politeness normally make it
awkward to issue flat imperative sentences (e.g. Leave the room) or
explicit performatives (e.g. I order you to leave the room), and we there-
fore seek to find indirect means to our illocutionary ends (e.g. I wonder if
you would mind leaving the room). In directives, politeness is the chief
motivation for indirectness. (Searle 1975:64)
I hope I have shown that it is an illusion to think that 'ordinary conversa-
tional requirements of politeness make it awkward to issue flat impera-
tive sentences'. It is not an 'ordinary' requirement, it is an English
requirement.
Similarly, the rule that 'it is awkward to issue explicit performatives'
is an English conversational requirement, not a universal one. The
awkwardness of the utterance quoted at the outset:
Please! Sit! Sit!
stems precisely from the fact that Polish does not share the two conversa-
tional requirements mentioned by Searle as 'ordinary'. (Please is a rough
equivalent of the Polish word which literally means I ask. In
P.olish speech, the performative form meaning 'I ask' is simply ubiqui-
tous, even more so than the highly colloquial form ci 'I advise
you'.) Furthermore, considering sentences such as:
Will you bloody well hurry up?
Why don't you shut your mouth?
one wonders how much explanatory force can be attributed to the claim
that 'politeness is the chief motivation for indirectness', even if one
limits this claim to English. (Cf. Ervin-Tripp 1976:59-61.)
From the data discussed in this chapter, it emerges that what is at issue
is neither universal rules of politeness nor even English-specific rules of
politeness. What is really at issue is English conversational strategies,
and Anglo-Saxon cultural values.
In an interesting study of politeness markers in English and German,
House - Kasper (1981:184) have observed that "on the whole, the
German speakers selected more direct levels for both complaint and
request acts". The authors comment on this difference as follows: "From
an etic standpoint, then, the behavior of the German speakers may well
Theoretical implications 61
be considered impolite by reference to an English norm; however, from
an emic standpoint, which is the one we would prefer here, one would
simply claim that the differential behavior displayed by the German and
English speakers may be a reflection of the fact that the two cultural
systems are organised differently, and that, for example, a level 6 com-
plaint in the German culture is not necessarily comparable to a level 6
complaint in the English culture, because the value of each is derived
from the value it has relative to the remaining levels, and their frequency
and modality of use in the particular cultural system." (1981: 184).
To my mind, however, the significance of the differences observed
goes much deeper than that. What is at issue is not just different ways
of expressing politeness, but different cultural values. As I see it, the
crucial fact is that different pragmatic norms reflect different hierarchies
of values characteristic of different cultures.
Commenting on the form Can you, Searle (1975:74-75) says: "Firstly,
X does not presume to know about Y's abilities, as he would if he
issued an imperative sentence; and, secondly, the form gives - or at
least appears to give - Y the option of refusing, since a yes-no ques-
tion allows no as a possible answer. Hence, compliance can be made to
appear a free act rather than obeying a command."
This is all true and insightful. It is an illusion, however, to think that
the norms referred to in this passage have the same weight in all cultures.
Searle is not unaware that "there are differences in the indirect speech
forms from one language to another", but he regards such differences
as idiomatic, due to accidental variation (1975:76). He explains:
The mechanisms are not peculiar to this language or that, but at the same
time the standard forms from one language will not always maintain their
indirect speech potential when translated from one language to another. ...
within the class of idiomatic sentences, some forms tend to become
entrenched as conventional devices for indirect speech acts. In the case of
directives, in which politeness is the chief motivation for the indirect
forms, certain forms are conventionally used as polite requests. Which
kinds of forms are selected will, in all likelihood, vary from one language
to another. (Searle 1975:76-77)
But this makes it sound as if the variation were more or less random
and accidental, whereas the general mechanisms were universal.
In fact, as I have tried to show, specific differences between languages
in the area of so-called 'indirect' speech acts are motivated, to a consid-
erable degree, by differences in cultural norms and cultural assumptions,
62 Different cultures, different languages, different speech acts
and the general mechanisms themselves are culture-specific. (Cf. Hollos
- Beeman 1978:353-354.)
This is not to deny that the generalisations suggested in works such
as Grice (1975), Gordon - Lakoff (1975) or Searle (1975) provide
useful insights into mechanisms of language use. It is important, how-
ever, that generalisations of this kind should not be seen as absolute.
'Natural logic' provides a considerable range of options. The choices
embodied in individual languages reflect not only 'natural logic' , and not
only a combination of 'natural logic' with historical accidents. They
reflect also what Gumperz (1982:182) aptly calls 'cultural logic'. Searle
insists that interrogative English sentences such as:
Can you pass the salt(?)
Would you pass me the salt(?)
Will you pass me the salt(?)
are not ambiguous (between question and request), but that by virtue of
their meaning they are simply questions (even when they are uttered with
intonation characteristic of directives, cf. Searle 1975:69). If they are
interpreted as requests, that is by virtue of the hearers' "general powers
of rationality and inference" (Searle 1979: 176).
But to say this is to imply that speakers of languages such as Polish
are sadly lacking those 'powers of rationality and inference'. Poles
learning English must be taught the potential ambiguity of Would you
sentences, or Why don't you sentences, just as they must be taught the
polysemy of the word bank. Searle might say that what they have to be
taught is not meaning but 'conventions of usage' (cf. Searle 1975:76).
But this distinction between meaning and conventions of usage becomes
meaningless if the ignorance of the relevant 'conventions of usage' leads
not just to un-idiomatic speech but to simple misunderstanding of what
Searle himself would recognise as meaning. For example, if Polish
newcomers to Australia interpret sentences such as:
How about a beer?
Why don',t you come and have lunch with us?
as genuine questions, rather than as an offer and an invitation, they are
making a semantic error just as much as when they interpret the utterance
How do you do? as a genuine question.
It is essential to recognise that what is involved is not any differences
in 'powers of rationality and inference', but differences in 'cultural
logic', encoded in language:
Theoretical implications 63
The fact that two speakers whose sentences are quite grammatical can
differ radically in their interpretation of each other's verbal strategies
indicates that conversational management does rest on linguistic knowl-
edge. But to find out what that knowledge is we must abandon the existing
views of communication which draw a basic distinction between cultural
or social knowledge on the one hand and linguistic signalling processes on
the other. We cannot regard meaning as the output of non-linear process-
ing in which sounds are mapped into morphemes, clauses and sentences
by application of the grammatical and semantic rules of sentence-level
linguistic analysis, and look at social norms as extralinguistic forces which
merely determine how and under what conditions such meaning units are
used. (Gumperz 1982:185-186)
I would add that descriptions of 'cultural logic', to be helpful, must be
done in fairly specific terms. It is worth noting in this connection that
in numerous studies written by Western scholars and concerning non-
Western cultures epithets such as 'direct' or 'blunt' are used to refer to
the Anglo-Saxon cultural norms, whereas, by contrast, the other cultures
studied often appear to value 'indirectness' (cf. for example Geertz 1976;
Eades 1982). In the present study, the reverse is the case: by comparison
with Polish, the English ways of speaking appear to be highly 'indirect'.
This shows, however, that terms such as 'directness' or 'indirectness' are
much too general, much too vague to be really safe in cross-cultural
studies, unless the specific nature of a given cultural norm is spelt out.
The present study shows that English cultural norms (as compared
with Polish norms) favour 'indirectness' in acts aiming at bringing about
an action from the addressee. On the other hand, studies such as Eades
(1982), Sansom (1980) or Abrahams (1976) show that Anglo-Saxon
cultural norms (as compared with Australian Aboriginal norms, or with
Black American norms) encourage 'directness' in seeking information
from the addressee. Evidently, the Anglo-Saxon principle of non-inter-
ference, which accounts for the heavy restrictions on the use of the
imperative, doesn't extend to questions (I don't mean 'personal ques-
tions', but questions in general) - presumably, because information is
seen in Anglo-Saxon culture as a free and public good. In fact, the
restrictions on the use of the imperative seem to be compensated by a
tremendous expansion of interrogative devices.
Similarly, Geertz (1976:240-248) stresses the 'indirection' and 'dis-
simulation' characteristic of Javanese culture, and contrasts these
features with those characteristic of American culture. According to
Geertz's classical study, Javanese culture favours "beating about the
64 Different cultures, different languages, different speech acts
bush", "not saying what is on one's mind', "unwillingness to face issues
in their naked truth", "never saying what one really thinks", avoiding
"gratuitous truths", "never showing one's real feelings directly" and so
on. Clearly, all these forms of 'indirection' are rather different from
those cultivated in Anglo-Saxon culture (especially, the dissimulation
of truth).
It seems to me, therefore, that it is very important to try to link
language-specific norms of interaction with specific cultural values,
such as autonomy of the individual and anti-dogmaticism in Anglo-
Saxon culture or cordiality and warmth in Polish culture. The issues
involved are of fundamental importance, and they merit a more general
discussion; I attempt to undertake such a discussion in Chapter 3.
6. Practical implications
In a multi-ethnic country like Australia, or like the United States, the
problem of speech acts and of their cultural significance is not a purely
academic one. It is a problem of immense practical significance.
As long as it is widely assumed that English conversational routines
reflect what is 'ordinary', 'normal', 'natural' and 'logical', the prospects
for cultural understanding between immigrants and the Anglo-Saxon
population are not particularly bright. Anglo-Saxon institutions such as
schools, courts or government departments, as well as the streets and
'market places' are, inevitably, an arena of cultural clashes and cultural
misunderstandings. If immigrants who speak passable English tend to
utter flat imperatives, they are likely to be seen as rude or boorish. If
they fail to respond to pieces of elaborate 'indirection', they are likely
to be seen as uncooperative, or dumb. Elaborate indirectness accompa-
nied by juicy swearing can be as confusing to an immigrant as the
directness, forcefulness and 'emotionality' of some immigrants can be
offensive and irritating to an 'Anglo'.
Anglo-Saxon doctors and nurses (as Jane Simpson has pointed out
to me) are accustomed to thinking that pain should be borne stoically,
and that one should only cry in real extremity. Therefore they are unsym-
pathetic to people who complain, cry and scream at pains which can be
considered minor, behaviour acceptable to Italians and Greeks. This can
lead to very unsympathetic treatment by doctors and nurses, and to a
Practical implications 65
general idea that Mediterranean peoples are cowardly because they
complain about things that only hysterical cowardly Anglo-Saxons
would mention. I have heard similar comments from Australian nurses,
during two seminars on linguistic problems of immigrants which I
gave to nurses in two Canberra hospitals in 1983. A number of
nurses commented on the unsympathetic attitude of Anglo-Saxon
doctors towards immigrant women screaming in childbirth, and on the
fact that often injections are administered merely to stop the screaming.
An immigrant woman who screams, cries or complains, is seen as
hysterical or unbalanced. The taboo on showing pain is clearly related to
the taboo on showing emotions.
Obviously, cultural clashes of this kind cannot be completely
eliminated, but they can be minimised by enlightened, well-planned
multicultural education. It seems clear that a linguistic study of culture-
specific speech acts and speech styles has a great deal to contribute in
this domain.
Chapter 3
Cross-cultural pragmatics and different
cultural values
Anyone who has lived for a long time in two different countries knows
that in different countries people speak in different ways - not only
because they use different linguistic codes, involving different lexicons
and different grammars, but also because their ways of using the codes
are different. Some of these differences are so stable and so systematic
that one cannot always draw a line between different codes and different
ways of using the code; or between different 'grammars' and different
'ethnographies of speaking' (cf. Hymes 1962).
The extent of the differences between different societies and different
language communities in their ways of speaking is often underestimated
in the literature dealing with language use. In particular, theories of
speech acts and of conversational logic associated with, or following
from, the work of philosophers such as John Searle (1969, 1979) and
Paul Grice (1975, 1981) have tended to assume that the ways of speak-
ing characteristic of mainstream white American English represent 'the
normal human ways of speaking', and that, apart from minor variations,
they can be expected to be the same as those prevalent in any other
human society. But this is of course an ethnocentric illusion.
The search for universals in language usage at the expense of culture-
specifics is also a feature of the influential study of 'politeness phenom-
ena' by Brown - Levinson (1978; revised edition 1987). There would
of course be nothing wrong in focussing on universals rather than on
culture-specific aspects of language usage - if the search for universals
is undertaken from a truly universalist, culture-independent position. But
as a number of recent studies have shown, the basic conceptual tools
introduced and relied on by Brown and Levinson (in particular, the
notion of 'face') have in fact a strong anglocentric bias (cf. for example
Matsumoto 1988; Katriel 1986; Tannen 1984; Wierzbicka 1985a, b).
Brown - Levinson see two principles as the most important ones
in human interaction: 'avoidance of imposition' ('negative face') and
'approval of the other person', which they exemplify with the English
compliment What lovely roses! ('positive face'). But their very choice
68 Cross-cultural pragmatics and different cultural values
of these particular parameters reflects clearly the authors' culture-
specific (anglocentric) perspective.
The same charge of anglocentrism can be made with respect to various
other supposedly universal 'maxims' and principles of human conversa-
tional behaviour and interaction, which have been advanced in the litera-
ture. Consider, for example, Leech's (1983:132) maxims of 'modesty'
and of 'approbation':
Approbation maxim
(a) Minimise dispraise of other; [(b) Maximise praise of other.]
Modesty maxim
(a) Minimise praise of self; [(b) Maximise dispraise of self.]
Leech is aware that the weight of maxims such as these may vary from
culture to culture, but he assumes that apart from quantitative differ-
ences they are in essence universally valid. In fact, however, empirical
evidence suggests that this is simply not true.
For example, Kochman (1981) has shown that in Black American
culture the norm of 'modesty' does not apply, and that self-praise is not
viewed negatively at all. Kochman mentions in this connection the title
of Mohammed Ali's autobiography: I am the greatest, and he discusses
the significance of Black folk categories such as 'rapping', 'grandstand-
ing', and 'showboating' (I return to this matter in section 1.5 below).
Similarly, Mizutani - Mizutani (1987) show that 'approbation' or
'praise of other' is not encouraged in Japanese culture; and they devote a
whole section (1987:45-46) to "refraining from direct praise". Likewise,
Honna - Hoffer (1989:74) point out that 'praise of other' is seen as
arrogant and presumptuous in Japanese culture, where "even when [the
speaker] has to or wants to express his praise for persons within his
circle, he often begins with a phrase such as 'I don't really mean to
praise ... ' or 'I know it is too presumptuous to praise ... '. By so doing
he tries to give the impression that he is not really an arrogant person."
It is not true, then, that all human societies view 'praise of self'
negatively, and 'praise of other' positively.
The same applies to the supposedly universal maxims of harmony:
"minimise disagreement, maximise agreement" (Leech 1983: 132). For
example, as Schiffrin (1984) has shown, Jewish culture displays a clear
preference for disagreement: in this culture, people show their involve-
ment with other people and their interest in other people by saying 'no'
rather than 'yes'. In Jewish culture, argument is valued as a form of
Cross-cultural pragmatics and different cultural values 69
sociability, and it is disagreement rather than agreement that is seen
as something that brings people closer together (see section 2.1 below).
It is, then, an anglocentric illusion to think that all cultures value
agreement more than disagreement, discourage self-praise, encourage
praise of other, and view 'imposition' as the main sin in social inter-
action.
The last decade has witnessed a growing reaction against this kind
of misguided universalism, a reaction which has led to the emergence
of a new field and a new direction in language studies associated with
the term 'cross-cultural pragmatics' (cf. for example Abrahams 1976;
Ameka 1987; Eades 1982; Goddard 1985; Harkins 1988; Hijirida -
Sohn 1986; Katriel 1986; Kochman 1981; Mizutani - Mizutani 1987;
Ochs 1976; Schiffrin 1984; Sohn 1983; Tannen 1981a; Wierzbicka
1985a, b). The main ideas which have informed and illuminated this
new direction in the study of language are these:
(1) In different societies, and different communities, people
speak differently.
(2) These differences in ways of speaking are profound and
systematic.
(3) These differences reflect different cultural values, or at
least different hierarchies of values.
(4) Different ways of speaking, different communicative
styles, can be explained and made sense of, in terms of
independently established different cultural values and
cultural priorities.
These four points are, in my view, of fundamental importance - not
only from the point of view of our knowledge and understanding of the
world, but also from a practical, social point of view; and in particular,
from the point of view of cross-cultural understanding in a multi-ethnic
society such as the United States or Australia.
Consider, for example, the situation of Australians of Anglo-Saxon
or Anglo-Celtic background who note that some immigrants behave
verbally in what appear to be strange, unfamiliar ways. For example,
they seem to shout and scream for no reason at all, they interrupt other
people, they start heated arguments for no apparent reason, they speak
in what is perceived as a blunt, dogmatic and bossy way, they flatly
assert their opinions and flatly contradict other people, and so on.
If 'strange' and possibly offensive behaviour of this kind can be ex-
plained, and made sense of, in terms of independently understandable
70 Cross-cultural pragmatics and different cultural values
cultural values, serious social and interpersonal problems can be
resolved, and serious conflicts prevented or alleviated. Of course, not
all problems can be solved in this way: if there is a real conflict in
underlying values, mere explaining will not help. But in many cases,
perhaps in most cases, what is involved is not a real conflict in values
but a difference in the hierarchy of values; and when this is the case,
explaining can help.
It can only help, however, if it is done in a way which is intelligible
to the target audience. And this is, I believe, where cross-cultural prag-
matics often fails. Even the most enlightened studies in cross-cultural
pragmatics (such as for example Kochman 1981; Sohn 1983; Lebra
1976) tend to explain different cultural priorities associated with differ-
ent languages (or different dialects and sociolects) in ways which are
not, and which cannot be, comprehensible to people of different cultural
backgrounds. The crux of the matter lies in the language in which the
explanations are couched.
What usually happens is that researchers in cross-cultural pragmatics
try to explain differences in the ways of speaking in terms of values such
as 'directness' or 'indirectness', 'solidarity', 'spontaneity', 'sincerity',
'social harmony', 'cordiality', 'self-assertion', 'intimacy', 'self-expres-
sion', and so on, without explaining what they mean by these terms, and
using them as if they were self-explanatory. But if one compares the
ways in which different writers use these terms, it becomes obvious that
they don't mean the same things for everyone. In fact, the intended
meanings are often not only different but mutually incompatible. As
a result, the same ways of speaking are described by some authors as
'direct' and by others as 'indirect'; as a manifestation of 'self-assertion'
or an absence of 'self-assertion'; as an expression of individuality or
suppression of individuality. This leads to total confusion, and to an
absence of any consensus, even on the most basic points.
For example, in the literature on Japanese culture and society,
Japanese ways of speaking are often described as 'indirect' and are
contrasted with the English ways of speaking, which are supposed to be
more 'direct'. It is also claimed, or even assumed, that English ways
of speaking are characterised by a high degree of self-assertion,
whereas in Japanese self-assertion is avoided and suppressed. It is
also said that English ways of speaking reflect high regard for
sincerity and spontaneity, whereas Japanese ways of speaking discourage
sincerity and spontaneity, preferring to them courtesy and consideration
for others.
Cross-cultural pragmatics and different cultural values 71
On the other hand, in the literature on (American) Black English, the
'standard' (white) English is presented in the opposite way. Here, it is
said, and even assumed, that standard English is 'indirect' rather than
'direct', that it avoids self-assertion, and that it discourages sincerity
and spontaneity. It is Black English which is said to be 'direct', and to
favour self-assertion, sincerity and spontaneity. Similarly, in the litera-
ture on Jewish culture, on the Yiddish language and also on Israeli
Hebrew, Yiddish and Hebrew are presented as 'direct', as bent on self-
expression and self-assertion, and as favouring sincerity and spontaneity,
whereas English is presented as associated with the suppression of all
these values.
At first, one might think that conflicting assertions of this kind are due
simply to differences of degree: perhaps English (that is, standard white
English) is more 'direct' or more 'self-assertive' than Japanese but less
so than Black English or than Israeli Hebrew. But when one examines
the data adduced in support of the conflicting generalisations, one
discovers that this is not the case, and that in fact the differences referred
to are qualitative rather than quantitative. For example, what is called
'self-assertion' in the studies of Black English is not the same thing
that is usually meant by this term in the studies of Japanese; and the
same applies to 'self-expression', 'sincerity', 'spontaneity', 'solidarity'
and so on.
I conclude from this that labels of this kind are simply not helpful
in the elucidation of cultural differences. Labels of this kind are semi-
technical and obscure at the same time. They are used differently by
different writers because they have no clear or self-evident meaning.
They are also highly anglocentric, as they have no exact equivalents
in other languages. For example, Japanese has no words corresponding
to sincerity. The two Japanese words which are usually translated
as 'sincerity', magokoro and makoto, mean in fact something very
different from sincerity, as Ruth Benedict (1947) among others has
clearly demonstrated. Nor does Japanese - or, for that matter, Polish,
Italian, French or Russian - have a word for self-assertion.
It seems obvious that if we want to compare "different cultures in terms
of their true basic values, and if we want to do it in a way that would
help us to understand those cultures, we should try to do it not in terms
of our own conceptual artefacts (such as the English terms self-assertion
or sincerity) but in terms of concepts which may be relevant to those
other cultures as well - that is, in terms of concepts which are
relatively, if not absolutely, universal. We should also try to do it in
72 Cross-cultural pragmatics and different cultural values
terms of concepts which are intuitively clear and intuitively verifiable,
and which t h ~ r o r will not be used differently by different scholars
and in different cultural contexts.
This may seem a tall order, but I submit that it can be done if we rely
on such simple and universal or near-universal concepts as want, say,
know, think, good and bad. In this chapter I shall try to demonstrate the
explanatory value of this approach by examining a number of parameters
which are widely relied on in the literature, seeking to clarify the sources
of confusion, and to reveal the real differences between languages
obscured by the use of confusing and inconsistently applied labels.
1. 'Self-assertion'
1.1. 'Self-assertion' in Japanese and in English
From a Japanese point of view, Western culture in general and Anglo-
American culture in particular can be seen as dominated by 'self-asser-
tion'. For example, Lebra (1976:257) contrasts "the Western model based
on the complex of individuality, autonomy, equality, rationality, aggres-
sion, and self-assertion" with "the traditional [Japanese] complex of
collectivism, interdependence, superordination-subordination, empathy,
sentimentality, introspection, and self-denial".
Similarly, Suzuki (1986) emphasises the Japanese tendency to avoid
'self-assertion' and the difficulties which this creates for the Japanese
in contact with Westerners:
We, used to assimilation and dependency, expect to project ourselves onto
the other, and expect him to empathise with us. We have great difficulty
with the idea that so long as our addressee is not Japanese we can't expect
to have our position understood without strong self-assertion. But estab-
lishing our own viewpoint or position before out addressee has understood
is not our forte ... So when Japanese, who aren't good at foreign languages,
don't show their true ability in international conferences and scholarly
meetings, it is less because of their language skills than because of the
weak development of the will to express themselves linguistically to suffi-
cient degree. It lies furthermore in the underdeveloped ability to stand
apart from the position taken by another and at least assert oneself to the
extent of saying, 'This is where I stand at this moment.' (Suzuki 1986: 157)
'Self-assertion' 73
On the other hand, when Kochman (1981:29) compares "the capaci-
ties and inclinations of whites and blacks [in America] to assert them-
selves" he sees the whites (that is, the members of the mainstream
Anglo-American culture) as less able, and less inclined, to assert
themselves. According to Kochman, "black culture allows its members
considerably greater freedom to assert and express themselves than
does white culture". He illustrates this claim, among other things, with
the different attitudes of white and black culture towards boasting and
bragging: "White boasting and bragging also contrasts with black prac-
tice with respect to the etiquette governing self-assertion. As white
culture restricts individual self-assertion generally, it requires that
individuals be governed by the norms of modesty when characterising
their performance" (1981 :69).
Thus, according to Kochman, white Anglo-American culture restricts
individual self-assertion, whereas accord,ing to Lebra or Suzuki, the
same white Anglo-American culture strongly encourages individual
self-assertion. Who is right and who is wrong? My view is that both
sides are right in what they are trying to say, but that they both fail to say
it clearly and unambiguously. Both sides use the same label 'self-asser-
tion', but they don't define it, and in fact they mean something quite
different by it.
The main difference between Japanese and mainstream English in the
area under discussion can be represented in terms of certain clearly
specifiable underlying conceptual structures. These structures are, above
all, these two:
Japanese
Anglo-American
don't say: 'I want this', 'I don't want this'
do say: 'I want this', 'I don't want this'
Japanese culture discourages people from saying clearly what they
want and what they don't want, whereas Anglo-Saxon culture, on the
contrary, encourages them to do so.
In a similar vein, Japanese culture discourages people from expressing
clearly their wishes, their preferences, and their desires (what they
would or wouldn't like or want), whereas Anglo-Saxon culture encour-
ages them to do so:
Japanese
Anglo-American
don't say: 'I would/wouldn't like (want) this'
do say: 'I would/wouldn't like (want) this'
74 Cross-cultural pragmatics and different cultural values
Furthermore, Japanese culture, in contrast to Anglo-American culture,
discourages clear and unequivocal expression of personal opinions;
Japanese
Anglo-American
don't say: 'I think this / 1 don't think this'
do say: 'I think this / 1 don't think this'
As pointed out by Smith (1983:44-45), "the Japanese are at pains to
avoid contention and confrontation ... much of the definition of a 'good
person' involves restraint in the expression of personal desires and
opinions". This restraint manifests one of the greatest Japanese cultural
values, called enryo, a word usually translated as 'restraint' or 'reserve'.
"One way to express enryo is to avoid giving opinions and to sidestep
choices when they are offered. As a matter of fact, choices are less
often offered in Japan than in the United States." (Smith 1983:83-84)
Smith quotes in this connection Japanese psychiatrist Takeo Doi' s
account of the strain he experienced on a visit to the United States,
where he was constantly offered choices:
Another thing that made me nervous was the custom whereby an American
host will ask a guest, before a meal, whether he would prefer a strong or a
soft drink. Then, if the guest asks for liquor, he will ask him whether, for
example, he prefers scotch or bourbon. When the guest has made this
decision, he next has to give instructions as to how much he wishes to
drink, and how he wants it served. With the main meal, fortunately, one
has only to eat what one is served, but once it is over one has to choose
whether to take coffee or tea, and - in even greater detail - whether
one wants it with sugar, milk, and so on.... I could not have cared less.
(Doi 1973:12)
Smith comments:
The strain must have been considerable, for in Japan, by contrast, the host,
having carefully considered what is most likely to please this particular
guest, will simply place before him a succession of an overwhelming
number of items of food and drink, all of which he is urged to consume, in
the standard phrase, 'without enryo'. It is incumbent on the guest to eat and
drink at least part of everything offered him, whether or not he likes the
particular item, in order not to give offence by appearing to rebuke his
host for miscalculating what would please him. (Smith 1983:84)
Since Japanese culture places a taboo on direct expression of one's
wants, it is also culturally inappropriate to ask other people directly
what they want. Mizutani and Mizutani explain:
'Self-assertion' 75
Asking someone' s wishes directly is also impolite In Japan. Saying
things like
*Nani-o tabetai-desu-ka. (What do you want to eat?)
*Nani-ga hoshii-desu-ka. (What do you want to have?)
should be limited to one's family or close friends. ... To be polite, one
should ask for instructions rather than directly inquire into someone' s
wishes. Thus, saying:
Mado-o akemashoo-ka. (Shall I open the window?)
is more appropriate than
*Mado-o akete-moraita-desu-ka. (Would you like me to open the
window?)
(Mizutani - Mizutani 1987:49)
The same cultural constraint prevents people in Japan from clearly
stating their preferences, even in response to direct questions. Many
Japanese, when asked about their convenience, decline to state it, saying
instead, for example:
Itsu-demo kekkoo-desu. (Any time will do.)
Doko-demo kekkoo-desu. (Any place will be all right with me.)
Nan-demo kamaimasen. (Anything will be all right with me.)
(Mizutani - Mizutani 1987:117-118)
"In actuality one cannot always agree to what another person wishes, and
one will then have to state one's own convenience anyway, but it is
regarded as childish to immediately start stating one's own convenience
when asked." (1987:118)
What applies to the expression of one's wants applies also the expres-
sion of one's opinions. This, too, comes under the value of enryo. Lebra
(1976:29) writes: "Pressure for conformity often results in a type of self-
restraint called enryo, refraining from expressing disagreement with
whatever appears to be the majority opinion." But "the virtue of enryo,
'self-restraint', is exercised not only to respond to group pressure for
conformity but to avoid causing displeasure for others, regardless of
their group membership ... The imposition of self-restraint to avoid
hurting Alter's feelings ... can reach an extreme that reveals immaturity
even to most Japanese. The individual may acquiesce in the face of an
intrusion on his rights or autonomy only because he is reluctant to offend
another person by claiming his right." (Lebra 1976:41-42)
I believe that the English concept of self-assertion is just as confusing
and unhelpful when applied to Japanese culture, as the Japanese concept
of enryo would be if applied to Anglo-American culture. On the other
76 Cross-cultural pragmatics and different cultural values
hand, the concept of enryo provides an essential key to understanding
Japanese culture. But to be able to use this key, we must first understand
what this concept really means; and we cannot understand it by trying to
translate it into English cultural concepts such as reserve, restraint,
modesty, or self-effacement. We can only understand it if we translate
it into culture-independent, universal or near-universal concepts such
as want, think, say, good or bad. This can be done in the following way:
enryo
X thinks:
I can't say to this person: I want this, I don't want this
I think this, I don't think this
someone can feel something bad because of this
X doesn't say it because of this
X doesn't do some things because of this
Difficulties experienced by the Japanese in dealings with Americans
(of the kind described by Doi) highlight the fact that no similar value is
embodied in Anglo-American culture. On the contrary, in English one
is expected to say clearly and unequivocally what one wants, what one
would like, or what one thinks. If that is what is meant by 'self-asser-
tion', then uninhibited self-assertion is indeed allowed and encouraged
in mainstream Anglo-American culture - as long as it doesn't come into
conflict with another cherished value of the culture, that is, personal
autonomy. This means that while one is allowed to say, in principle, 'I
want X', one is not allowed to say freely:
I want you to do X
since in this case, the speaker's right to 'self-assertion' would come
into conflict with the addressee's right to personal autonomy. This is
why in English the use of the bare imperative is very limited, and why
directives tend to take an interrogative or semi-interrogative form in
English.
This means that in English there is a strong cultural constraint on
saying to other people something that would amount to 'I want you to do
X'. Instead, one is expected to combine this component with some other
components, which would recognise the addressee's personal autonomy,
for example:
'Self-assertion' 77
I want you to do X
I don't know if you will do it
I want you to say if you will do it
This or a similar combination of components can be realised in English
by means of interrogative-directive devices (sometimes called 'whim-
peratives ') such as:
Would you do X?
Will you do X?
Could you do X?
Can you do X?
Why don't you do X?
and so on. By contrast, in many other languages, for example Polish
(Chapter 2 above), Russian (Comrie 1984a), Hebrew (Blum-Kulka -
Olshtain 1984; Blum-Kulka - Danet - Gherson 1985), Italian (Bates
1976), and Hungarian (Hollos - Beeman 1978), the bare imperative
is used much more freely, and the use of interrogative structures in
directives is much more limited.
In fact, even in Japanese, the use of interrogative structures in direc-
tives is more limited than in English (see for example Matsumoto 1988).
This does not mean that Japanese encourages the use of the bare impera-
tive any more than English does. But in Japanese, the important thing
is to show deference and to acknowledge one's dependence on other
people rather than to avoid imposition. As Matsumoto (1988) rightly
points out, non-imposition based on individual rights is an Anglo-Saxon
(or Anglo-American), not a universal value. For example, in Japanese it
is very polite to start interaction with other people by uttering 'direct'
requests, such as
Doozo yoroshiku onegaisimasu.
(lit.) 'I ask you to please treat me well.'
Musume 0 doozo yoroshiku onegaisimasu.
(lit.) 'I ask you to please treat/take care of my daughter well.'
Matsumoto (1988:410) observes that in utterances of this kind the
speakers "in indicating that they, or someone closely related to them,
are someone who needs to be taken care of by the addressee, humble
themselves and place themselves in a lower position. This is certainly
typical of deferential behaviour. The speech act in question, however,
is a direct request; thus, an imposition. ... it is an honour to be asked to
78 Cross-cultural pragmatics and different cultural values
take care of someone in that it indicates that one is regarded as holding
a higher position in the society."
This means that, in many situations, it is easier to say 'I want you
to do X' than 'I want to do X' - as long as one acknowledges one's
dependence on the addressee:
I want you to do X
I know that you don't have to do it
I say: it will be good for me if you do it
I think: you will do it because of this
In English, if one wants the addressee to do something, it is important
to acknowledge the addressee's autonomy by inviting them to say
whether or not they will comply with the request. Hence, the prolifera-
tion and the frequency of 'whimperatives' in English. In Japanese, inter-
rogative directive devices or 'whimperatives' exist, too, but their scope
is much narrower than in English (cf. Matsumoto 1988; Kageyama -
Tamori 1976). Instead, there is in Japanese a proliferation of devices
acknowledging dependence on other people, and deference to other
people. Hence, the basic way of making requests in Japanese involves
not 'whimperatives' (i.e. quasi-interrogative structures) but dependence-
acknowledging devices (usually combined with expressions of respect):
V-te kudasai.
'Give me (please) the favour of doing V.'
'I feel respect towards you.'
Even speaking to a child one would usually phrase a request in terms
of 'favours', although the expression of respect would be omitted:
V-te kure.
'Give me (please) the favour of doing V.'
1.2. 'Self-assertion' in black and white American English
When we turn to the comparisons between what have been called black
and white speech styles in America, we see that the term 'self-assertion'
stands here for rather different features of verbal behaviour than those
to which it usually refers in the literature contrasting English with
Japanese. For example, Kochman writes:
'Self-assertion' 79
Black culture values individually regulated self-assertion. It also values
spontaneous expression of feeling. As a result, black cultural events typi-
cally encourage and even require individuals to behave in an assertive/
expressive manner, as in such black speech events as rapping and signify-
ing ... and, as I am claiming here, argument. (Kochman 1981:29-30)
Similarly, when white American culture is described in terms of 'self-
restraint', this word doesn't stand for the same thing for which it stands
in the literature on Japanese culture. Another example from Kochman:
White culture values the ability of individuals to rein in their impulses.
White cultural events do not allow for individually initiated self-assertion
or the spontaneous expression of feeling. Rather, self-assertion occurs
as a social entitlement, a prerogative of one's higher status or, as with
turn-taking, something granted and regulated by an empowered authority.
And even when granted, it is a low-keyed assertion, showing detachment,
modesty, understatement. ... 'Showing off', which would represent
individually initiated (unauthorised) self-assertion and more unrestrained
self-expression, is viewed negatively within white culture. Black culture,
on the other hand, views showing off - in black idiom stylin' out,
showboating, grandstanding - positively.... Because white culture
requires that individuals check those impulses that come from within,
whites become able practitioners of self-restraint. However, this practice
has an inhibiting effect on their ability to be spontaneously self-assertive.
(Kochman 1981:30)
Clearly, 'the ability of individuals to rein in their impulses' is some-
thing quite different from the ability to say clearly what one thinks, what
one wants to do, or what one's preferences are. If the Japanese 'self-
restraint' consists mainly in refraining from saying 'I want X', the white
Anglo-American 'self-restraint' consists largely in refraining from say-
ing now what I want now and from saying what I think the moment I
think it.
The very principle of turn-taking, regarded as fundamental in Anglo-
American culture, forces the individual speakers to 'rein in' their
impulses to some extent. In black culture - as in Jewish culture (cf.
Tannen 1981b) - different speakers are allowed to speak all at once, to
overlap with one another and to interrupt one another, to share in this
way excitement, interest, and mutual involvement, and to maintain a
continuous flow of uninhibited communication and self-expression. But
this is not a difference between saying and not saying 'I want X'. Rather,
it is a difference between saying it at once and saying it at what one
sees as an appropriate moment.
80 Cross-cultural pragmatics and different cultural values
1.3. Spontaneity, autonomy, and turn-taking: English vs. Japanese
This doesn't mean that white Anglo-American culture generally discour-
ages spontaneous self-expression. Rather, it discourages it to the extent
to which spontaneous self-expression might come into conflict with the
principle of everyone's personal autonomy: one can express oneself
spontaneously, if this doesn't infringe on what is seen as other people's
right to speak without interruptions and without interference from
other people.
It is interesting to note in this connection that in the literature compar-
ing Japanese culture with mainstream Anglo-American culture, the latter
is usually said to encourage rather than discourage 'spontaneity'. For
example, the author of a study comparing Japanese and American
educational materials writes:
Is the expression of spontaneous feelings encouraged or discouraged? ...
Japanese teachers are advised to discourage students from expressing
impulsive thought and emotional opinions ... The result may be the estab-
lishing of two identities, one functioning on the communicative level
and the other known only to ego. A damper is based on the potential for
shared excitement. There are, of course, inhibitors in the United States.
The difference is a matter of degree. (Lanham 1986:294)
But I don't think it is a matter of degree. Rather, it is a matter of
different cultural priorities. In Japanese culture, the overriding cultural
principle seems to be constant caution not to offend or not to hurt other
people (and also to avoid embarrassment for oneself which could
follow from this); that is, an attitude which can be portrayed as follows:
if I do X someone could feel something bad because of this
I don't want this
The Anglo-American principle of personal autonomy can be represented
as follows:
everyone can say: 'I want this', 'I don't want this'
'I think this', 'I don't think this'
one can't say to someone: 'you have to do X because I want it'
'you can't do X because I don't
want it'
The Anglo-American principle of tum-taking can be seen as a manifesta-
tion of this more general principle of personal autonomy, and of a more
'Self-assertion' 81
general respect for the rights of every individual. The principle of
tum-taking can be represented as follows:
someone is saying something now
I can't say something at the same time
I can say something after this
It is interesting to note that Japanese culture doesn't observe the same
principle of turn-taking. On the contrary, since Japanese culture values
interdependence more highly than autonomy, in Japanese conversation
utterances are expected to be, to a large extent, a collective work of the
speaker and the addressee, or, more generally, of different speakers. This
is done, in particular, by means of 'response words', that is, of what is
called in Japanese aizuchi, a word which likens Japanese conversation to
the work of two swordsmiths hammering a blade in tum. Mizutani -
Mizutani (1987:18-19) write:"The word ai means 'doing something
together' ... ; tsuchi means 'a hammer' .... Two people talking and
frequently exchanging response words is thus likened to the way two
swordsmiths hammer on a blade. In Japanese conversation, the listener
constantly helps the speaker with aizuchi ... - the roles of the speaker
and the listener are not completely separated." Mizutani and Mizutani
stress that aizuchi are absolutely essential to Japanese conversation and
they support this with a startling statistic: 'The average number of
aizuchi per minute is ... from 12 to 26, according to the study made by
one of the authors." (1987:20)
This is a striking manifestation of the Japanese value of interdepend-
ence, which is just the opposite of the Anglo-American principle of
personal autonomy. The same applies to the Japanese conversational
principle of leaving sentences unfinished so that the addressee can com-
plete them. As Mizutani - Mizutani (1987:27) describe it, in Japanese,
"leaving a part of the sentence unsaid so that the listener can supplement
it is often more considerate and polite than just going ahead and complet-
ing one's own sentence.... always completing one's own sentences can
sound as if one is refusing to let the other person participate in complet-
ing a sentence which might better be completed by two people".
The attitude reflected in Japanese conversational style can be por-
trayed as follows:
I want to say something now
I think you know what I want to say
I think you would say the same
82 Cross-cultural pragmatics and different cultural values
I think I can say part of it, you can say another part of it
I think this will be good
Thus, if the Anglo-American conversational principle of turn-taking
reflects the cultural value of personal autonomy, the Japanese conversa-
tional principle of 'collective sentence production' reflects the Japanese
cultural values of interdependence, co-operation, and 'groupism'.
1.4. 'Spontaneous self-assertion' vs. 'regulated self-assertion':
black English vs. white English vs. Japanese
Returning now to black English, we note that although it too rejects the
tum-taking model, it doesn't reject it in favour of the conversational
co-operation and interdependence characteristic of Japanese. On the con-
trary, it rejects it in favour of what Kochman (1981) calls spontaneous
or impulsive self-assertion and self-expression, that is to say, in favour
of some values which are contrary to the Japanese ethos. It is interesting
to note that Kochman describes the contrast between black English
and white English in this respect using the same pair of terms that, for
example, Barnlund (1975b:35) uses to describe the contrast between
white English and Japanese: 'regulated' vs. 'spontaneous'. Thus, accord-
ing to Kochman, black English is 'spontaneous' and white English
'regulated', whereas according to Bamlund, Japanese is 'regulated' and
English (that is to say, white English) is 'spontaneous'. But this means
that the same white English that from a Japanese perspective is seen
as 'spontaneous' and 'not regulated', is seen from a black perspective
as 'regulated' and 'not spontaneous'.
The 'regulated' character of white English means, roughly speaking,
that while one can express one's thoughts, wants, and feelings, one is
expected to observe certain rules in doing so; in particular, one is ex-
pected not to interrupt other people, and not to speak at the same time
as other people. This constrains one's spontaneity, to some extent, but
it doesn't constrain one's freedom of self-expression.
On the other hand, in Japanese one is expected to be much more
circumspect in expressing one's thoughts, one's wants, and one's feel-
ings. It is not only a question of when to express them, but whether
one should express them at all; Japanese discourse can be said to be
'regulated' with respect to what to say, not just when to say it. When the
Japanese self is described as a "guarded self' (for example by Barnlund
'Self-assertion' 83
1975b: 112), reference is made, in the first place, to what is said, and in
particular, to the care Japanese speakers take "to prevent overexposure
of inner selves" (1975b:112). Bamlund illustrates this 'restraint' in self-
disclosure with striking statistical data, showing enormous differences
between Americans and Japanese in the range of topic they are prepared
to talk about, and also in the range of persons to whom they are prepared
to reveal their thoughts and their opinions.
As for when, the important thing is not so much not to overlap with
other people, as to premeditate what one is going to say in order to
avoid saying something which could hurt or offend somebody, or
which could embarrass the speaker him/herself. Thus, Barnlund
(1975b:131) describes Japanese communication as "a three-act play:
'Premeditation', 'Rehearsal', and 'Performance"'. One can see why the
terms 'regulated' and 'non-spontaneous' can come to mind in this
connection, but clearly this cannot be the same thing as Kochman has
in mind when he describes white American English as 'regulated' and
'non-spontaneous'. This shows, once again, that labels such as 'regu-
lated' or 'spontaneous' are not self-explanatory, just as ' self-assertion'
and 'self-expression' are not self-explanatory, and are used by different
writers to apply to different phenomena, and to different cultural norms.
On the other hand, semantic formulae couched in terms of universal
semantic primitives can be both precise and self-explanatory. I propose
the following:
Black American culture
I want/think/feel something now
I want to say it ('self-assertion', 'self-expression')
I want to say it now ('spontaneity')
White Anglo-American culture
I want/think/feel something
I want to say it ('self-assertion', 'self-expression')
I cannot say it now
because someone else is saying something now ('autonomy',
'tum-taking')
Japanese culture
I can't say: I want/I think/I feel something
someone could feel something bad because of this
if I want to say something
I have to think about it before I say it
84 Cross-cultural pragmatics and different cultural values
1.5. 'Self-assertion' as personal display: black English vs. white
English
But the alleged 'self-assertiveness', attributed by Kochman and others to
American black culture, has other features, which are reflected in char-
acteristic black styles and genres such as sty/in' out, showboating, and
grannin' (grandstanding) (Kochman 1981). Each of these concepts de-
serves detailed analysis, which cannot be undertaken here. All that I can
do in the present context is to point out to some characteristic cultural
features which are manifested in these and other similar folk-concepts.
The black so-called self-assertion consists largely in an uninhibited
desire to draw attention to oneself, and to behave, verbally and non-
verbally, in ways which would ensure this. As a first approximation,
this can be represented as follows:
I want people to think about me now
I want to do something because of this now
In addition to this general desire for attention, however, there is also the
more specific desire for admiration, or rather, for admiring attention - a
desire which in black culture is viewed positively, not negatively. This
is clearly visible, for example, in black boasting, bragging, and overt
exultation and jubilation over one's success. For example, Kochman
(1981 :72) cites a television interview with some black basketball play-
ers, who had just won a championship basketball game. "One of the main
players of the team, asked to comment on their opponents, was serious at
first, talking about 'playing hard and matching us height for height', etc.
However, he ended up with the exultant and self-congratulatory 'But we
were just too good for them!' "
As a first approximation, we could portray this attitude as follows:
I know:
I can do good things
other people can't do the same
I feel something good because of this
I want people to think good things about me because of this
It is important to recognise, however, that in black culture self-aggran-
disement of this kind has a somewhat theatrical quality, and that it is
meant partly as public entertainment. To reflect this vital aspect of black
'self-aggrandisement' one important component has to be added to the
formula sketched above:
'Self-assertion' 85
I know:
I can do good things
other people can't do the same
I feel something good because of this
I want people to think good things about me because of this
I say this because I want people to feel something good
Kochman (1981:73) points out that in black culture, boasting is inter-
preted "not as an unwarranted and uncouth claim to superiority but as
humour"; or as Reisman (1974:60) puts it, as "the assertion of oneself,
the making of one's noise, which depends not so much on the specific
content of the boast as on the fact that it is made -loudly - at all". The
expression 'assertion of oneself' appears here again, but, again, the con-
text makes it clear that it is not the same 'assertion of oneself' which the
literature on Japanese language and culture attributes to mainstream
Anglo-American culture.
1.6. 'Self-assertion' and 'good interpersonal relations'
One might hypothesise that all cultures cherish and seek to promote
'good relations' among people. But different cultures interpret this goal
differently, and they seek to implement it in different ways; and these
different interpretations are reflected in different 'ethnographies of
speaking'. In Japanese culture, the prevailing conceptual formula is this:
if I do/say something someone could feel something bad because
of this
I don't want this
I have to think about it before I do it
This is why Japanese culture can be seen as a 'culture of anticipatory
perception' and a 'culture of consideration' (Suzuki 1986:157), a culture
bent on preventing displeasure. Lebra (1976:41) remarks: "One should
note how often in speech the Japanese refer to the need not to cause
meiwaku, 'trouble', for another person, not to be in his way, and not to
hurt his feelings. In actual behaviour, too, they tend to be circumspect
and reserved, so as not to offend other people."
In black American culture there is no similar emphasis on preventing
displeasure, and, consequently, there is no emphasis on 'self-restraint'.
On the contrary, black culture encourages uninhibited spontaneous self-
86 Cross-cultural pragmatics and different cultural values
expression. At the same time, however, it is a culture where self-expres-
sion and self-display is seen as conducive to 'good feelings', not only in
the speaker but also in other people; and as a means to promote shared
excitement, shared fun, shared interest, and shared 'colour'. A black
English derogatory term for white people is 'grey': white people are seen
as 'grey' not only because of the colour of their skin, but because of
what is perceived as their 'lifelessness', their 'moderate impassioned
behaviour', their lack of spontaneous emotionality, their 'reining in of
their impulses'. (Johnson 1972:144-145)
In white Anglo-American culture, the main emphasis is not on pre-
venting displeasure, or on spontaneous and uninhibited self-expression,
or on generating good feelings among one's 'audience', but on personal
autonomy (for everyone), on non-imposition, and non-interference. It is
a culture which encourages everyone to say freely - at the right time -
what they want and what they think, and (in a characteristic phrase) to
'agree to disagree'.
Thus, while one can presume that all cultures cherish and seek to
promote 'good relations' among people, it is not true that, for example,
both American culture and Japanese culture cherish 'warm and cordial'
relations among people, as asserted, for example, by Lanham (1986:293).
A cultural emphasis on interpersonal warmth (in private relations) can be
said to be characteristic of Russian culture (cf. for example Smith 1983),
but not of American or Japanese culture. Such an emphasis is reflected,
for example, in the extraordinary wealth of Russian expressive deriva-
tion, and in particular, in the abundance of hypocoristic forms of Russian
names (see Wierzbicka, to appear, chap. 7).
Japanese culture can be said to encourage empathy, consideration, and
avoidance of hurting others, but not warmth or cordiality, as is shown by
the extraordinary wealth, and wide use, of devices encoding 'apologies',
'quasi-apologies', 'preventive apologies', 'grateful apologies', and so on
(see for example Mizutani - Mizutani 1987; Coulmas 1981). The vir-
tual absence of linguistic devices encoding 'warmth' (in sharp contrast
with the wealth of devices encoding 'respect', 'deference', and the like),
points in the same direction. The relatively small degree of physical
contact and physical intimacy between people in Japanese society
provides further evidence for this (cf. Barnlund 1975b:106-108).
American culture encourages a generalised friendly attitude to people,
including strangers. But this, too, is different from the personalised af-
fection displayed, for example, in Russian hypocoristic names. The
American generalised friendliness can be seen in the common phrase
'Self-assertion' 87
Have a nice day! (often addressed to complete strangers, sometimes
even displayed on badges on the uniforms of shop assistants, or on
taxi windows).
These three different cultural emphases in the interpretation of 'good
interpersonal relations' can be represented as follows:
'Warmth', of the kind associated with Russian or Polish culture:
I feel something good towards you
'Considerateness' and empathy, of the kind associated with
Japanese culture:
I don't want someone to feel something bad
'General friendliness', of the kind associated with American
culture:
I want everyone to feel something good
Needless to say, the formulae sketched above are not meant to capture all
the different aspects of different cultural attitudes to emotions. For
example, for Japanese culture we might also posit the following rule:
I don't want to say what I feel
whereas for Russian or Polish culture we might postulate the opposite
norm:
I want to say what I feel
On the other hand, it would not be justified to posit for Japanese culture
the rule which seems to prevail in Javanese society, especially among the
Javanese gentry (prijaji):
I don't want people to know what I feel
For example, Geertz (1976:247) writes of the Javanese: "One often hears
people say in praise of someone that 'one can never tell how he feels
inside by how he behaves on the outside' "; and he speaks of "the nearly
absolute requirement never to show one's feelings directly, especially to
a guest" (1976:246). (See section 2.4 below.)
In Japan the norm seems to be different: not 'I should conceal what I
feel' but 'I should not verbalise what I feel'; that is, not 'I don't want
people to know what I feel' but 'I don't want to say what I feel'. The
whole Japanese emphasis on empathy, on omoiyari (cf. Lebra 1976:38-
49) shows that Japanese culture does not discourage an interest in other
people's emotions; quite the contrary. But it does discourage verbal
88 Cross-cultural pragmatics and different cultural values
expression of emotions. We could formulate, therefore, the following,
fuller, set of Japanese cultural norms related to emotions:
(1) I don't want someone to feel something bad
(2) I don't want to say what I feel
(3) I should know what this person can feel
this person doesn't have to say it
2. 'Directness'
The terms 'directness' and 'indirectness' are often used in linguistic
descriptions as if they were self-explanatory. In fact, however, they are
applied to totally different phenomena, which are shaped by totally
different values.
The confusion which surrounds this notion is linked with the widely
accepted distinction between so-called 'direct' and 'indirect' speech acts,
and in particular, between imperatives and the so-called whimperatives.
Thus, it is widely assumed that if one says to somebody Close the door!
this is a 'direct' speech act, whereas if one says Could you close the
door? or Would you mind closing the door? this is an 'indirect' speech
act. But although these particular examples may seem clear, it is by no
means clear how the distinction in question should be applied to other
phenomena and to other languages. Thus, in many languages, for ex-
ample, in Russian, Polish, Thai, or Japanese, the imperative is often
combined with various particles, some of them somewhat impatient,
others rather friendly, some of them described as 'softening' the direc-
tive, others as, on the contrary, making it harsher or more peremptory,
and so on. Are such combinations of the imperative with a particle
'direct' or 'indirect' speech acts? There is no general principle which
would allow us to answer this question.
I suggest, therefore, that the whole distinction between 'direct' and
'indirect' speech acts should be abandoned - at least until some clear
definition of these terms is provided; and also, that the distinction be-
tween 'direct' and 'indirect' ways of speaking in general should be
abandoned, and that the different phenomena associated with these
labels should be individually examined. I believe that when this is done,
the confusion surrounding these concepts can be cleared, and some clear
'Directness' 89
cultural explanations for the cross-linguistic differences associated with
these terms can be provided.
2.1. American culture vs. Israeli culture
According to Blum-Kulka - Danet - Gherson (1985: 133), "viewed
from a cross-cultural perspective, the general level of directness in
Israeli society is probably relatively very high". What exactly is meant
by this 'high level' of 'directness'? One clear example is provided by
the wide use of bare imperatives in social interaction, including public
interaction:
(Passenger to driver: on the bus)
Passenger A: ptax et hadelet, nehag
(Open the door, driver.) (No response.)
Passenger B: nexag, delet axorit.
(Driver, rear door.)
(Compliance.)
(Blum-Kulka - Danet - Gherson 1985:129)
Presumably, in English, an interrogative-directive device (could you or
would you) would be used in a similar situation, and the authors appear
to regard this as a clear case of directness vs. indirectness. In asking
directions from a stranger on the street, the standard procedure for
English is an 'attention-getter' (Excuse me... ) and the form Can/could
you tell me ... ? (Blum-Kulka 1982:46). But in Hebrew, the standard
procedure is a 'direct request for information' ('Where is the railway
station?').
What does 'directness' mean in cases of this kind? I think it means
that in Hebrew one can say rather freely something that means:
I want you to do (say) X
whereas in English, generally speaking, one is not expected to say
this without at the same time acknowledging the addressee's personal
autonomy:
I want you to do X
I don't know if you will do it
90 Cross-cultural pragmatics and different cultural values
Hence the combination of the imperative with some interrogative
features in common English directives.
Why should the American and the Israeli cultures differ in this way?
According to the above-mentioned source:
One possible explanation for this high level of directness is an ideological-
historical one: The early settlers of Palestine were guided by an ideology
of egalitarism, which frowned on all manifestations of possible discrimina-
tion between people, including a show of deference in speech. ... It is
against this background that one should consider the directness of present-
day Israeli society ... (Blum-Kulka - Danet - Gherson 1985:133-134)
But this explanation is hardly convincing, given the egalitarian ethos of
North America: surely, American culture doesn't encourage manifesta-
tions of discrimination among people, or 'a show of deference', either
(cf. de Tocqueville 1953).
The same authors (1985: 137) also offer another explanation: "... these
findings can be interpreted as reflecting the distinct, culture-specific
interactional style of Israeli society. The low value attached to social
distance, manifested in language by a relatively high level of directness,
suggests that the interactional style of this society is basically solidarity
politeness oriented." I think that this observation is more to the point
in comparing Hebrew with English, but, unfortunately, terms such as
'social distance' or 'solidarity politeness' are not self-explanatory either.
Trying to really understand the cultural values in question, we could
propose for Israeli Hebrew the following formula:
we can all say to one another: 'I want you to do this'
we will not feel something bad towards one another because of
this
Since in Israeli Hebrew one can also freely express one's 'diswants'
(for example, in refusals, disagreements, and so on) the formula above
should probably be expanded so as to include 'I don't want' as well as
'I want'. For example, Blum-Kulka observes in an earlier work:
Generally speaking, Israeli society seems to allow for even more directness
in social interaction than the American one .... It is not uncommon to hear
people around a conference table in Israel disagreeing with each other
bluntly (saying things like ata to'e 'You're wrong', or 10 naxon! 'Not
true! '). Such directness in a similar setting in American society would
probably be considered rude. Similarly, refusal is often expressed in Israel
by a curt ' No'; the same 10 (no) can also be heard as a response to requests
iDirectness' 91
phrased as requests for information ('Do you have such and such?') in
shops, hotels, and restaurants, a habit that probably contributes to the
popular view about Israelis' 'lack of politeness'. (Blum-Kulka 1982:30-31)
We can portray the Israeli attitudes in question as follows:
we can all say to one another:
'I want this', 'I don't want this', 'I think this', 'I don't think
this'
we will not feel something bad (towards one another?) because
of this
In Anglo-American culture, too, one can say fairly freely what one
wants, what one doesn't want, and what one thinks, but one is not
expected to be similarly 'blunt' about it, because it is as important in this
culture to acknowledge everyone's right to independence and personal
autonomy as to exercise one's own right to self-expression. Furthermore,
in Anglo-American culture there is no emphasis on 'we' (corresponding
to the cultural value of 'solidarity' in Israeli culture); rather, there is a
strong emphasis on every individual's separate and autonomous 'I'. This
is sometimes described in terms of 'rugged individualism' as opposed to
an 'ethos of solidarity' (cf. for example Arensberg - Niehoff 1975); but
there are many ways to be 'individualistic' and many ways to be 'non-
individualistic' or 'anti-individualistic'. For example, the Israeli ethos of
'solidarity' (cf. Katriel 1986) is different from, though related to, the
Australian ethos of 'mateship' (cf. Wierzbicka 1986b); and it is certainly
different from the Japanese ethos of 'dependence' and 'groupism' (cf.
for example Lebra 1976; Smith 1983). Here as elsewhere, therefore,
in spelling out cultural values it is safer to rely on explicit semantic
formulae than on undefined and protean global labels such as 'direct-
ness', 'individualism', 'solidarity' or 'collectivism'. We can portray the
Anglo-American cultural assumption in question as follows:
I think: I can say: 'I want this', 'I think this'
I know: other people don't have to want the same/think the same
no one can say: 'I want you to want this', 'I want you to think
this'
I have not included in this formula the component 'I don't want this',
because Anglo-American culture does impose certain inhibitions on
the expression of 'diswants' and doesn't encourage open confrontation.
In Hebrew, and in Jewish tradition in general, open confrontation is
92 Cross-cultural pragmatics and different cultural values
encouraged and cherished, as a reflection of spontaneity, closeness and
mutual trust. As the Jewish writer Sholom Aleichem put it (quoted in
Myerhoff 1978:188): "We fight to keep warm. That's how we survive."
(cf. also Schriffrin 1984). In Anglo-American culture, however, 'direct
confrontation' is avoided in the interests of social harmony between
independent individuals. In view of the emphasis on individualism and
on everyone's personal autonomy, 'closeness' is cherished in this culture
less than 'harmony'.
In saying this, I am contradicting the view of many Japanese scholars,
who see Japanese culture as a culture of 'harmony' and Anglo-American
culture as one which positively encourages 'direct contention and con-
frontation'. But this just shows, once again, that global labels such as
'harmony' are used by different writers in different senses. The fact of
the matter is that, as pointed out by Blum-Kulka (1982:30-31) or by
Levenston (1970), in England or in America it is not common to hear
people around a conference table disagree with one another by saying
'you're wrong' or 'that's not true'; in fact, it is not common to use such
phrases in everyday conversation either. Anglo-American tradition
encourages people to say 'I don't think so' rather than 'you are wrong'.
Japanese culture discourages people even from saying 'I don't think so'.
But we cannot accurately account for all such differences in terms of
labels such as 'harmony', 'directness', or 'confrontation'.
Blum-Kulka (1982:30-31) mentions that it is not common in English
to express refusal by saying 'No' as one does in Hebrew, or to say 'No'
in response to a request for information (for example in shops, hotels,
and restaurants): 'Do you have such and such?'. In English, when some-
one indicates that they want something from us we are free to say 'No',
but not to say just 'No'. The label 'directness' is not helpful in describ-
ing this aspect of the Anglo-American ethnography of speaking, though
one can use here, more illuminatingly, the label 'bluntness'. (It should be
noted, however, that 'bluntness', though clearer here than 'directness',
is not self-explanatory either, and that for example Geertz (1976:245)
attributes 'bluntness' to Anglo-American culture, contrasting it in this
way with Javanese culture.) 'Bluntness' in saying 'no' is viewed posi-
tively in Israeli culture but not in Anglo-American culture. These differ-
ent attitudes to 'bluntness' in saying 'No' can be represented as follows:
Anglo-American culture
I say: No
'Directness' 93
I don't want you to feel something bad because of this
I will say something more about it because of this
Israeli culture
I say: No
I think I don't have to say anything more about it
In Japanese culture, the norm seems to be to avoid saying 'No' alto-
gether (in particular, to refuse an offer or a request, to express disagree-
ment, and so on). Thus, Nakane (1970:35) notes: " ... one would prefer
to be silent than utter such words as 'no' or 'I disagree'. The avoidance
of such open and bald negative expressions is rooted in the fear that
it might disrupt the harmony and order of the group". This norm can be
represented as follows:
Japanese culture
I can't say: No
I will say something else because of this
Barnlund (1975b) explicitly compares the Japanese with the Americans
in this respect:
Anyone who has observed groups of Japanese or Americans talking
together is aware at once of certain peculiarities in their habits of speech.
In one group everyone bows and exchanges personal cards. When they
speak they do so quietly, often in the form of understatements. Rarely does
one hear a belligerent or unequivocal 'No.' ... In the other group, they all
shake hands as they begin a conversation. 'No' is heard at least as often
or more often than ' Yes'. ... Arguments are heated, issues often polarised.
(Barnlund 1975b:26-27)
But if the difference between the Americans and the Japanese is pre-
sented in such a polarised manner, it is hard to see how the same Ameri-
cans can appear to the Israelis as people who, in contrast to themselves,
avoid saying 'No'. It seems to me that the semantic formulae proposed
here allow us to paint a clearer and more coherent overall picture.
2.2. 'Indirectness' in Japanese
According to Mizutani - Mizutani (1987), Honna - Hoffer (1989), and
many other writers on Japanese language and culture, it is extremely
important when talking politely in Japanese 'to sound indirect'. But what
does one do in Japanese 'to sound indirect'?
94 Cross-cultural pragmatics and different cultural values
First of all, one doesn't say what one wants; instead, one sends
'implicit messages', expecting that the addressee will respond to them:
The speaker thus often makes indirect requests, and the listener also
responds to implicit messages: this makes the indirect development of
speech possible. For instance, a man, usually a superior, will come into
the room and say:
Kyoo-wa iya-ni atsui-nee. (It's awfully hot today, isn't it?)
And one of his men will say hai ['yes', respectful], and hurry to open the
window or turn on the air conditioner. He may even apologise saying:
Doomo ki-ga tsukimasen-de ... (I'm sorry I didn't notice.)
... many Japanese seem to find pleasure in being with someone who under-
stands them very well and so will sense their wishes and act to realise them
without being asked.
(Mizutani - Mizutani 1987:36)
The attitude manifested in speech behaviour of this kind can be repre-
sented as follows:
I want something
I don't want to say this
I will say something else because of this
I think this person will know what I want
A different phenomenon, also described in the literature in terms of
'indirectness', has to do with deliberate lack of precision and lack of
specificity in the identification of referents, or in using numbers:
In social situations the Japanese like to refer to numbers or amounts in a
non-specific way. For instance, when buying apples they will often say:
Mittsu-hodo/gurai/bakari kudasai. (Please give me about three of them.)
instead of saying
Mittsu kudasai.
(Mizutani - Mizutani 1987:33)
Furthermore, in making proposals or suggestions, the Japanese tend to
refer to things with indirect expressions like demo and nado (and others).
For example:
Ocha-demo nomimasen-ka.
'How about having some tea?' [lit. 'or something?']
Eiga-demo mimashoo-ka.
{Directness' 95
'How about going to a movie?' [lit. 'How about seeing a movie
or something?']
(Mizutani - Mizutani 1987:34)
Similarly:
A: Mada jikan-ga aru-n-desu kedo.
'I have some time to kill.'
B: la, zasshi-demo yondara doo-desu-ka.
'Then, why don't you read a magazine or something?'
(Mizutani - Mizutani 1987:34)
In such situations, ocha-demo or eega-demo are preferred to ocha-o or
eega-o because they let the listener choose among several possibilities.
This deliberate use of non-specific reference and non-specific numeral
expressions can be portrayed as follows:
I say: I would want something like this
I don't want to say: 'I want this'
It is not difficult to recognise here again the Japanese value of enryo,
discussed earlier - a value which is quite different from the Anglo-
American value of personal autonomy. But if all the different phenom-
ena in question are described by means of the same label 'indirectness'
then the different cultural values involved cannot be revealed, and
the generalisations made in individual works devoted to comparisons
of two cultures do not seem to make sense in a broader cross-cultural
perspective.
As a particularly striking example of the resulting confusion I now
tum to the question of 'which culture encourages more "indirectness" -
Greek or American?'
2.3. Greek culture and American culture
Consider first the following statement, fairly characteristic of the way
the concept of 'indirectness' tends to be used in the literature on cross-
cultural pragmatics:
Though languages provide their speakers with explicit, direct ways for
achieving communicative ends, in day-to-day communication speakers
seem to prefer indirect ways. In making a request to a secretary, for
96 Cross-cultural pragmatics and different cultural values
example, people are more likely to say things like 'Could you do it' or
'Would you mind doing it' than the simple 'Do it'. (Blum-Kulka 1982:30)
The writer of the above knows very well that the generalisation in ques-
tion applies not to 'people in general' but mainly to Anglo-Saxons, and
that, for example, it doesn't apply to the Israelis. But this doesn't prevent
her from formulating it as if it in fact applied to 'people in general'.
Furthermore, the illustration provided makes it clear that what the
author has in mind is the phenomenon of 'whimperatives', directives
phrased interrogatively; but the generalisation is couched in terms of
'indirect ways of speaking' - as if it were enough to mention the
'whimperatives' to explain what one means by 'indirect ways of speak-
ing' in general.
Blum-Kulka (1982:30) proceeds then to make the important and, I
think, perfectly valid point that "one major factor that can influence the
application of such principles can be the general 'ethos' of one society
as compared to another one". But having said this, she says something
rather startling, that is, that "Greek social norms, for example (Tannen
[1981a]), require a much higher level of indirectness in social interaction
than American ones" (Blum-Kulka 1982:30).
This statement might lead one to believe that if in Israel one tends to
say 'Do it!' more widely than one does in America, in America one
tends to say 'Do it!' more widely than one does in Greece; and that,
conversely, if in America one tends to say 'Would you' or 'Could you'
in many situations in which in Israel one would say simply 'Do it!', in
Greece one tends to say 'Would you' or 'Could you' in many situations
in which in America one would say simply 'Do it!'.
But is this believable? Surely not. In fact, a claim of this kind would
seem to go against everything one knows about Mediterranean culture
generally, and about the Greek culture more specifically. In particular,
the characterisation of Greek culture as 'indirect' or as 'more indirect'
than American culture, seems to be incompatible with the results of
behavioural studies devoted specifically to the Greek national character,
and of behavioural differences between Greeks and Americans, such
as Triandis - Vassiliou (1972). For example, according to this study,
typical Greek behaviour shows characteristics that an American will
interpret as arrogance, dogmatism, and attempts to appear all-knowing
and all-powerful.
'Directness' 97
The characterisation of Greek culture as 'indirect' also goes against
the expectation that Greece and Middle East (including Israel) might
share some cultural values, and some features of their ethnography of
speaking (cf. Tannen - Oztek 1977; Matisoff 1979) rather than being
at the opposite poles of a scale, with Anglo-Saxon ways of speaking
in the middle:
'direct'
'intermediate'
'indirect'
Israel
England and North America
Greece
One can only wonder where Japan would appear on a scale of this kind?
Below Greece, perhaps? And (American) black English? Above Israel?
I believe that here as elsewhere, scales are misleading and confusing
if they are not preceeded by rigorous qualitative analysis. If one exam-
ines the data in Blum-Kulka's source of information on Greek culture
(Tannen 1981a), it transpires that the so-called Greek 'indirectness'
applies to phenomena quite different from the use of whimperatives; and
the whole puzzling story of 'Greek indirectness' versus 'American
directness' begins to make sense.
What Tannen did was to present a number of informants (some
Americans, some Greeks, and some Greek-Americans) with a written
questionnaire, which begins by presenting an exchange between a wife
and a husband:
Wife: John's having a party. Wanna go?
Husband: Okay.
Two paraphrases are then presented, and respondents are asked to indi-
cate which they believe the husband meant when he said okay:
(1-I) My wife wants to go to this party, since she asked.
I'll go to make her happy. ['indirect']
(I-D) My wife is asking if I want to go to a party.
I feel like going, so I'll say yes. ['direct']
Tannen's results are clear and interesting: "A comparison of the percent-
age of respondents in the three groups who opted for paraphrase 1-1 turns
out looking much like a continuum, with Greeks the most likely to
take the indirect interpretation, Americans the least likely, and Greek-
Americans in the middle, somewhat closer to Greeks." (Tannen
1981a:229).
98 Cross-cultural pragmatics and different cultural values
Although Tannen herself describes her study as dealing with 'modes
of indirectness', she is generally careful to point out that she is dealing
only with one specific context: a negotiation between husband and wife
about whether to go to a party. Nonetheless, some of her comments
could be seen as inviting the kind of over-generalisation expressed in
Blum-Kulka's account of her study. For example, she reports that "an
American-born woman of Greek grandparents ... commented that she
tends to be indirect because she picked it up from her mother, who was
influenced by her own mother (i.e. the grandmother born in Greece)"
(Tannen 1981a:235). Similarly, she quotes another personal testimony
which she calls "most eloquent": "that of a professional man living in
New York City, whose grandparents were from Greece. He seemed fully
assimilated, did not speak Greek, had not been raised in a Greek neigh-
bourhood, and had few Greek friends. In filling out the questionnaire, he
chose 1-1, the initial indirect interpretation. In later discussion he said
that the notion of indirectness 'rang such a bell'." (1981a:235)
This really could lead one to believe that Greek culture is somehow
generally 'indirect', certainly more so than American culture. But what
does this really mean? All that Tannen has really shown is that Greek
couples seem to be more attuned to one another's unexpressed wishes
than American couples are, and more ready to guess one another's unex-
pressed wishes, whereas American couples seem to rely more on explicit
verbalisations of wishes. In fact, some of Tannen's comments suggest
that in Greek culture it is the woman who is generally expected to guess,
and to comply with, her father's, or her husband's, unexpressed wishes:
For example, a Greek woman of about 65 told me that before she had
married she had to ask her father's permission before doing anything.
She noted that of course he never explicitly denied her permission. If
she asked, for example, whether or not she should go to a dance, and
he answered,
(1) An thes, pas. ('If you want, you can go.')
she knew that she could not go. If he really meant that she could go, he
would say,
(2) Ne. Na pas. ('Yes. You should go.')
... This informant added that her husband responds to her requests in the
same way. She therefore agrees to do what he prefers without expecting
him to express his preference directly. (Tannen 1981a:224-225)
But if this is all there is to it, is it enough to draw the conclusion that
"Greek social norms ... require a much higher level of indirectness in
social interaction than American ones" (Blum-Kulka 1982:30)? It seems
'Directness' 99
to me that a conclusion of this kind is unwarranted and misleading. On
the other hand, Tannen's data suggest the following cultural norm,
which seems to be quite credible, clear, and meaningful:
I want something
I don't have to say this
I think this person will know what I want
I think she will do it because of this
It is particularly interesting to note here the difference between the
Japanese general enryo ('reserve, self-restraint'):
I want something
I don't want to say this
and the Greek (male, typically) self-confidence:
I want something
I don't have to say this
(I think she will do it anyway)
It is also relevant to mention the importance of the division between
'in-group' and 'out-group' in Greek culture, and the great intimacy
and closeness prevailing within the 'in-group'. Triandis - Vassiliou
(1972:304) speak in this connection of the existence of an "extremely
tightly knit family and an 'ingroup' that provides protection, social
insurance, and a warm and relaxing environment; in short, a haven
from the larger world". In a warm, intimate environment of this kind
one doesn't have to rely on overt, verbal expression of one's needs,
wishes and desires.
As for Anglo-American culture, Tannen's findings are perfectly
consistent with the general Anglo-American emphasis on everyone's
personal autonomy and on the individualism prevailing even within the
family: Anglo-American culture encourages people to say, clearly and
explicitly, what they want and what they think. Apparently, American
spouses, too, rely less on wordless communication, and more on clear
self-expression. Possibly, this implies less of a feeling of 'oneness'
between the spouses, and a greater emphasis on each spouse's individu-
ality, unpredictability, and personal autonomy. All this is consistent with
what we otherwise know of Anglo-American cultural values. The term
'indirectness' doesn't really help us here. In fact, it is rather an obstacle
to understanding.
100 Cross-cultural pragmatics and different cultural values
2.4. 'Indirectness' and 'dissimulation' in Javanese
According to Geertz (1976:244), indirectness or 'indirection' is a major
theme of Javanese behaviour. Geertz illustrates this feature with the
proverb 'to look north and hit south'. He also mentions the fact that
old-time kijajis (Koranic teachers) never explicitly informed people
they were wrong, but told little stories from which the listeners could get
the point less painfully. "One must get the rasa of what people are
saying, the real content, informants are always emphasising, because
alus people (i.e. civilised people) often don't like to say what is on
their minds."
'Indirectness' as described above is closely related to another
Javanese cultural norm, that is, to what Geertz calls 'dissimulation or
pretence', or what the Javanese themselves call efok-efok. "The charac-
teristic quality of efok-efok, in contrast to our patterns of dissemblance,
is not merely that it is far more prevalent and that it is largely approved
... but that it need not have any obvious justification, being merely
gratuitous.... In general, polite Javanese avoid gratuitous truths."
(1976:245-246). Thus, Geertz quotes the following definition of efok-
efok, offered by an informant:
He said: Suppose I go off south and you see me go. Later my son asks
you: 'Do you know where my father went?' And you say no, e{ok-e{ok you
don't know. I asked him why should I e{ok-efok, as there seemed to be no
reason for lying, and he said, 'Dh, you just e{ok-efok. You don't have to
have a reason.' (Geertz 1976:246)
This general cultural norm of concealment, of not saying, not telling
people any 'gratuitous truths', applies in particular to the truth about
one's personal feelings:
The same sort of pattern is involved in the nearly absolute requirement
never to show one's real feelings directly, especially to a guest. Any
kind of negative feeling towards another must be dissimulated.... Strong
positive feelings are also supposed to be hidden except in very intimate
situations. The effort is to keep a steady level of very mild positive affect
in interpersonal relations, an e{ok-e!ok warmth behind which all real
feelings can be effectively concealed. (Geertz 1976:246)
What applies to feelings applies also to wishes: one should conceal
one's wishes and one's intentions, particularly if they are in conflict
with other people's wishes or desires. For example:
tDirectness' 101
... one must call out to any passerby one knows inviting him to stop in,
even though he may be the last person on earth you wish to see. One must
refuse food (unless the host persists in offering it) even if one is dying of
hunger ... One should never refuse outright people's requests to do some-
thing for them; rather, one merely agrees even if one has no intention of
going through with whatever it is, and then one never gets around to doing
it, putting the petitioner off with various etok-etok excuses, until he realises
at last that one was not serious in the first place. (Geertz 1976:246-247)
Apparently, what applies to feelings and to wishes, applies also to
thoughts. Geertz (1976:247) quotes a village politician on this point, who
began his speech as follows: "No one ever says what he really thinks.
People always e{ok-erok when dealing with other people. 1 too never
say what 1 really think, and you can't tell how 1 feel about things by
what 1 say."
The reluctance to express one's feelings, wants, and thoughts links
Javanese culture with Japanese cultural norms described earlier; but
the element of concealment, of conscious 'dissimulation', seems to be
specifically Javanese. We can portray this 'dissimulation' as follows:
1 don't want to say: 1 feel X/I want X/I think X/I know X
1 don't want people to know what 1 feel/want/think/know
The more specific norm proscribing explicit requests can be portrayed
as follows:
1 can't say to someone: 'I want you to do X'
someone could feel something bad because of this
1 have to say something else
The norm proscribing explicit refusals can be portrayed along similar
lines:
if someone says to me:'1 want you to do X'
1 can't say: 'I don't want to do it'
someone could feel something bad because of this
1 have to say something else
1 don't have to do it because of this
The avoidance of providing 'gratuitous information' can be represented
as follows:
if someone says to me:
'you know something'
102 Cross-cultural pragmatics and different cultural values
'I want you to say it'
I can't say: 'I don't want to do it'
I can say something else
I don't have to do it because of this
The general principle of e!ok-e{ok can perhaps be formulated as follows:
I don't want to say what I think/know
I don't have to say this
I can say something else
In Western culture, saying what one thinks tends to be seen as
everyone's right, and saying what one knows, as everyone's obligation
(although there are of course limits to this). Generally speaking, then,
questions can be freely asked and answers cannot be freely withheld
(cf. Eades 1982 and the references quoted by her). These attitudes can
be portrayed as follows:
(1) I can say what I think
(2) I can say to people:
'you know something'
'I want to know it'
(I can think: they have to say it)
(3) if someone says to me:
'you know something'
'I want to know it'
I have to say it
In many non-Western cultures, however, and in particular in Javanese,
culture, a different norm prevails, which can be portrayed as follows:
if someone says to me:
'you know something'
'I want to know it'
I don't have to say it
As shown by Eades (1982), in Australian Aboriginal culture one
wouldn't even assume that one has the right to ask; on the contrary, the
opposite norm prevails:
I can't say to people:
'you know something'
'I want to know it'
'Directness' 103
Many writers have tried to explain cultural differences of this kind,
pointing to different cultural attitudes to knowledge, questions, and in-
formation (cf. Eades 1982; Abrahams 1976; Sansom 1980; Keen 1978;
Harris 1984; Goody 1978). While accepting their explanations, I would
like to add to them an additional one: different cultural attitudes to truth.
European culture has traditionally placed a great premium not only on
'knowing' but also on saying what one knows, that is, what is knowable
(or true). Other cultures may value knowledge without valuing verbal
articulation of knowledge. For example, Japanese culture is said to value
intuitive knowledge and to mistrust verbalised, articulated knowledge
(cf. for example Bamlund 1975b; Lebra 1976). It is interesting to note in
this connection that while all languages appear to have a word corre-
sponding to know, many languages do not have a word corresponding to
true (cf. Hill 1985). Some languages have a word for something like
lying (to another person), without having a word like true which com-
bines in its meaning 'knowing' and 'saying' (that is, 'saying what one
can know'), without any reference to interpersonal relations, as in the
case of 'lying' (cf. Lutz 1985:73).
In fact, even in English the word truth didn't always have the imper-
sonal and objective ring which it has now. As Hughes (1988:61-62)
observes, "The central and fascinating point in the semantic history of
truth is that it evolves from being a private commitment to a publicly
assessed quality. The form of word even changes, so that troth, the
private form, can, by the proof of arms, be asserted above even the
claims of evidence or testimony, if need arises. (This mediaevalised
form of truth is, of course, virtually the opposite of the modem notion,
which is factual, demonstrable and essentially impersonal.)"
European culture, however, elevated the truth (first the private,
personal 'truth', and then the public, impersonal truth) to a particularly
high place among generally accepted ideals; and 'truth' can be seen as
opposed to both 'lying' and 'concealment', to both saying what one
knows is not true and not saying what one knows is true. The cultural
norms in question can be represented, roughly, as follows:
it is bad to say what is not true
it is good to say what is true
It might be added that modem Anglo-American culture appears to be
more 'pragmatic' in its attitude to truth than European culture. This is
reflected, for example, in the concept of 'a white lie', which doesn't
seem to have any equivalents in German, French, Italian, or Polish
104 Cross-cultural pragmatics and different cultural values
(see section 3.5 below). Cultural attitudes to conscientiousness,
punctuality, or reliability may indeed differ along the lines suggested
by Max Weber (very roughly, between Protestant northern Europe, plus
its American extension, and the Catholic rest, cf. Weber 1968); but the
attitudes to pragmatic, 'white' lies may be divided along rather different
lines, with, roughly speaking, continental Europe on one side of the
dividing line and the more 'pragmatic' Anglo-American culture on the
other. This modified, Anglo-American attitude to truth can be repre-
sented, very roughly, as follows:
it is usually bad to say what is not true
sometimes it is good to say what is not true
if nothing bad can happen to anyone because of this
As one early Anglo-Saxon put it: "Use not to lie, for that is unhonest;
speak not every truth, for that is unneedful; yes, in time and place, a
harmless lie is a great deal better than a hurtful truth." (Roger Ascham,
1550, quoted by Stevenson 1946:2058). But the norm discouraging not-
truth (whether in an absolute or in a modified, 'pragmatic' form) is by no
means universal. In particular, the Javanese principle of etok-etok allows
one both not to say what one knows is true and also to say what one
knows is not true. Perceived cultural advantages involved in such an
attitude may include 'tranquillity', 'harmony', smooth and peaceful inter-
personal relations ('I don't want to feel something bad', 'I don't want
someone to feel something bad'), and so on.
3. Further illustrations: same labels, different values
In this section, I discuss in a more summary way the use of five other
global labels, which are generally believed to stand for identifiable
cultural values, but which in fact are used to refer to different attitudes
and different ways of speaking. I try to uncover the real differences in
cultural values, concealed and obscured by such inconsistently and arbi-
trarily applied terms. The labels in question are: 'intimacy', 'closeness'
(contrasted with 'distance'), 'informality' (contrasted with 'formality'),
'harmony', and 'sincerity'.
Further illustrations: same labels, different values 105
3.1. 'Intimacy'
It is widely believed that different cultures differ in the importance they
give to "intimacy' as a social value. For example, according to Hijirida
- Sohn (1986:390) American culture gives this value a high priority,
whereas in Japanese and Korean culture other values (for example, re-
spect for rank and status) by far 'overrule' intimacy as a cultural norm.
The claim that in American English 'intimacy' overrules rank or social
status, whereas the opposite is true of Japanese and Korean, is perhaps
not hard to believe, even without any precise definition of 'intimacy'.
But when the authors make a more general claim, attributing to Ameri-
cans an "extreme sensitivity toward the intimacy variable" (1986:391),
we cannot go along with this without asking what exactly is meant by
'intimacy', and how this 'sensitivity to intimacy' is assessed.
In fact, in my own analysis of Anglo-American culture as compared
with Polish culture (cf. Wierzbicka 1985b; see also Chapter 2 above) or
with Russian culture (cf. Wierzbicka, to appear), I have reached conclu-
sions very different from those suggested by Hijirida and Sohn. From
a Polish, or Russian, point of view, Anglo-American culture is not
'sensitive to intimacy' at all. What, then, is 'intimacy'?
If we were to rely on the everyday meaning of the word intimacy (and
what else can we rely on?), we could define the concept as follows:
Intimacy refers to a readiness to reveal to some particular persons some
aspects of one's personality and of one's inner world that one conceals
from other people; a readiness based on personal trust and on personal
'good feelings'. This last proviso is necessary because although one
might disclose one's secret fears or worries to a doctor or to a psycho-
analyst this doesn't qualify as 'intimacy': to count as intimacy, self-
disclosure has to be based on an assumption of personal good feelings.
This can be represented as follows:
intimacy
X thinks: I feel something
I want to say it to someone
I can say it to Y
I feel something good towards Y
Y feels something good towards me
I can say it to Y because of this
I can't say it to other people
X says it to Y because of this
106 Cross-cultural pragmatics and different cultural values
Barnlund (1975b), who as we have seen has shown that Americans
are more prone to self-disclosure than the Japanese, has concluded from
this that the former value intimacy more than the latter:
The Americans will tend to cultivate physical as well as verbal intimacy.
Since the aim is to seek more complete expression of the inner self, Ameri-
cans may not only disclose more fully verbally, but may try to utilise as
many channels of communication as possible. For this reason they may
display greater physical animation and engage in a higher frequency of
physical contact during conversation. Touch, as one of the more intimate
forms of interaction, may be more encouraged and more accepted.
(Barnlund 1975b:38)
But even touch ceases to be 'intimate' if it is applied indiscriminately.
A handshake may indeed be more revealing than a bow, but it is not
necessarily more intimate. If intimacy could be reduced to self-disclo-
sure, the claim that Americans are more given to intimacy than the
Japanese could be sustained. But although intimacy is indeed related
to self-disclosure, it cannot be reduced to it. To count as intimacy,
self-disclosure has to be selective (in terms of the addressee), and this
selectiveness has to be based on personal affection.
In my view, a culture where one basic term of address, 'you', is used
indiscriminately to everyone, cannot be regarded as one which attaches
a great importance to the value of intimacy. If anything, it is extremely
difficult to be intimate in English, because of this universal 'you', that
is, because of the absence of any 'intimate' forms of address.
There are of course nicknames, and so-called affectionate nicknames
(for example Bob and Bobby for Robert, Kate and Katie for Katherine);
but are these truly instruments of intimacy? Hijirida - Sohn (1986:391)
think that they are. They write: "The tendency of Americans to upgrade
address forms (from FN to TLN) toward a person they are angry at,
structural differentiation of FN into FFN, Nn, and ANn in E[nglish] and
the productive use of them all reflect the extreme sensitivity Americans
have toward the intimacy variable."
But I don't think this is right. It is not 'intimate' in English to call
somebody John rather than Dr. Brown, and 'nicknames' such as Bob or
Tim are no more intimate than John. As for so-called affectionate nick-
names, such as Bobby or Timmy, they are not intimate but child-oriented;
they can be affectionate, but affection is not the same thing as intimacy,
particularly if it is an affection associated with the adult-child style
of interaction. (For fuller analysis of forms of address and names see
Wierzbicka, to appear, chaps. 7,8.)
Further illustrations: same labels, different values 107
I conclude that there is no linguistic evidence for the claim that
English is particularly sensitive to intimacy. On the contrary, English,
with its absence of any 'intimate' form of address, seems to be particu-
larly insensitive to it. On the other hand, there is massive evidence for
the importance of intimacy in Slavic languages. This evidence takes
the form, above all, of enormous differentiation of expressive forms of
personal names, such as, for example, Katja, Katen' ka, Katjusen' ka,
Katecka, Katik and so on for Katerfna, or Vanja, Vanecka, Vanjusa,
Vanjuska, Vanjusecka, and so on, for Ivan (see Wierzbicka, to appear).
With respect to Polish, one can argue that the value of intimacy is
even enhanced by the wide use of titles and other linguistic devices
keyed to rank and status, since this increases the differentiation of
personal relations. Hijirida - Sohn (1986:389) state that "both Japanese
and Koreans, being extremely status-conscious, are eager to give and
receive powerladen titles in daily interpersonal encounters", and they
link this with the low value of intimacy in Korea and Japan. But Poles,
too, are extremely status-conscious, and are eager to give and receive
titles in daily interpersonal encounters; they also value a degree of
formality and ritualised courtesy. At the same time, however, Poles place
a high value on intimacy, and the wide range of possibilities between,
say, Pani Professor ('Mrs Professor', with a third person form of the
verb) and various intimate forms of expressive derivation of names,
enhances the value of intimacy enjoyed with those special people with
whom one chooses to share it.
It is an illusion, then, to think that an egalitarian ethos, such as that
prevailing in Anglo-American culture, leads necessarily to an increase in
intimacy, or that a culture sensitive to status distinctions is necessarily
inimical to intimacy. Once again, terminological confusion leads here
to conceptual confusion, clearly visible, for example, in the following
passage from an otherwise subtle and insightful study:
Of the three societies under comparison, American is least sensitive to
power variables, as evidenced in both the patterns and usages of honorifics.
This seems to be due to their egalitarian value orientation. As a result,
solidarity variables like intimacy and casualness prevail, although
groupness [sic] solidarity is the last thing for Americans to give heed to
due probably to their strong individualistic value orientation. (Hijirida -
Sohn 1986:383)
American society is described here as one dominated by 'solidarity
variables like intimacy and casualness', although at the same time
108 Cross-cultural pragmatics and different cultural values
solidarity is said to be 'the last thing for Americans to give heed to'. This
is confusing and self-contradictory. We can clear this confusion if we
stop using undefined labels such as 'intimacy', 'solidarity' or 'casual-
ness', and start using instead precise and self-explanatory semantic
formulae couched in terms of universal semantic primitives.
3.2. 'Closeness'
Speaking metaphorically, intimacy implies closeness - another variable
which often comes up in discussions devoted to cross-cultural pragmat-
ics. But what is 'closeness' in interpersonal relations and how does one
assess it? Social psychology has developed various measures that can
be used to assess 'social distance' (cf. for example Triandis - Triandis
1960; Bogardus 1933), but these have to do with relations between
groups, not between individuals, and cannot be transferred to the study
of interpersonal relations.
The concept of 'distance' in interpersonal relations is heavily relied
on by a number of writers on linguistic pragmatics, and in particular, by
Brown - Levinson (1978); but it is never defined, and it is treated as
if it was self-explanatory. At best, it is elucidated by means of examples.
For example, Brown and Levinson assert that:
only D[istance] varies in the following two sentences:
(1) Excuse me, ltvould you by any chance have the time?
(2) Got the time, mate?
Our intuitions are that (1) would be used where (in S's perception)
S[peaker] and H[earer] were distant (strangers from different parts, say),
and (2) where Sand H were close (either known to each other, or percepti-
bly 'similar' in social terms). D, then, is the only variable in our formula
that changes from (1) to (2) ... (Brown Levinson 1978:85)
But this is baffling and unhelpful. Two university professors are pre-
sumably 'similar in social terms', but it doesn't follow from this that
they would be likely to exchange phrases like Got the time, mate? On the
other hand, two young male hitchhikers may well address one another
in this way even if they are 'strangers from different parts'.
Similarly baffling and unhelpful is the further claim that'D' is held
constant in the following utterances:
(3) Excuse me, sir, would it be all right if I smoke?
(4) Mind if I smoke? (Brown - Levinson 1978:85)
Further illustrations: same labels, different values 109
To justify this claim, Brown and Levinson appeal to their 'intuitions',
but this just shows how unreliable and idiosyncratic such 'intuitions'
about abstract and semi-technical concepts like 'distance' may be.
It seems to me that if we are to rely on the everyday use of the word
close (as applied to human relations) we would have to say that close-
ness has to do with interpersonal 'knowledge' as well as interpersonal
feelings: two people are said to be 'close' if they know one another
very well, and have 'good feelings' for one another. This is similar to
intimacy, but it is not the same thing. For example, a mother can be
said to be very 'close' to her daughter, but it would be a little odd to
say that a mother is 'intimate' with her daughter. A mother who is
'close' to her daughter knows a great deal about the daughter - about
her 'hidden' thoughts, fears, hopes, desires, and so on.
Speech doesn't seem essential to the idea of 'closeness', but mutual
knowledge, and the willingness to let one another know what is happen-
ing inside us, does seem to be essential. Sometimes, 'closeness' may
even reduce the need for verbal self-disclosure: if two people are very
'close' they may each know how the other person feels without overt
speech, by a kind of empathy. But not all 'empathy' manifests 'close-
ness'; 'closeness' being a permanent (long-term) feature of a relation-
ship, based on mutual good feelings.
Tentatively:
closeness ('X and Yare close to one another')
X and Y know: we feel something good towards one another
because of this each of them thinks of the other:
I want to know what this person feels/thinks/wants
I want this person to know what I feel/think/want
because of this, each of them can know what the other feels/
thinks/wants when other people can't
To let someone become close to us means to trust them enough, and to
feel enough affection (or 'good feelings') for them, to allow them to
know us really well - better than other people know us. This may be
seen as dangerous, because knowing us so well the other person will
probably be able to hurt us. There will also be more opportunities for
clashes, for mutual hurt, for open conflict. It may be safer not to get too
'close' - if one values peace, harmony, absence of conflict and absence
of mutual hurt.
110 Cross-cultural pragmatics and different cultural values
Not all cultures, therefore, encourage closeness, certainly not to the
same degree. For example, if I think something bad about you (for
example that you look awful, or that you have done something bad) I
have the option of telling you this or of concealing this thought from
you. If I do tell you you may be hurt or offended, but at least you
will know what I think, and you will know that I am interested in your
actions and your appearance. Telling you could promote our closeness.
Not telling you is more likely to promote harmony. In a situation like
this Polish culture, or Russian culture, would tend to opt for telling (that
is, for closeness), and Anglo-American culture, for not telling (that is,
for harmony).
Or suppose that I have done, or want to do, something that I think you
would disapprove of. Should I tell you or not? If I tell you, this will
promote our closeness, but it will disrupt our harmony and peacefulness.
You may feel something bad because of this, you may feel angry, you
may express your disapproval, and you may make me angry and upset.
If I don't tell you, there will be no ill-feeling, but we will not be close.
Again, in a situation like this, Polish culture, or Russian culture, would
probably opt for telling, and Anglo-American culture for not telling. The
attitude of a person who cherishes and seeks closeness with another
person can be portrayed as follows:
I want you to know what I feel/think/want
I know that you can feel something bad because of this
I know that I can feel something bad because of this
I want you to know it
because I know that you feel something good towards me
I think you know that I feel something good towards you
If one wanted to put a global label on this attitude, one might suggest
'self-disclosure', or 'openness', but this would be misleading. As we
have seen, Barnlund (1975a) interprets his findings concerning American
culture in terms of 'self-disclosure', but clearly, the attitude he is talking
about is quite different from that portrayed here. In the 'self-disclosure'
discussed by Barnlund, the stress is on 'self', on 'I', on saying what 1
think. In the 'closeness' discussed here the stress is on the relationship
between 'I' and 'you', on good feelings between 'I' and 'you', and on
a desire to continue and to promote a special relationship between us
two, even at the cost of hurt and conflict.
Further illustrations: same labels, different values 111
Needless to say the 'closeness' portrayed here has little to do with the
attitude reflected in utterances such as
Got the time, mate?
Mind if I smoke?
Utterances of this kind could be described as informal or casual (among
other things), but if they were said to reflect either 'intimacy' or 'close-
ness', one would have to say that the words 'intimacy' and 'closeness'
are being used in some technical sense, not in the everyday sense, and
that without clear definitions such use of these words obscures, rather
than clarifies, the attitudes involved.
3.3. 'Informality'
Informality is a cultural attitude which, as we have seen, is frequently
confused with intimacy or closeness. In Australia, when one rings a
travel agency, one will often hear a response including the travel
clerk's first name, for example:
American Express, Cathy speaking.
If one were to believe Hijirida - Sohn (1986) one might conclude that
this travel clerk expresses intimacy or closeness towards her customers.
I have argued, however, that intimacy involves a 'special relationship'
between two people, which certainly does not apply in the present
case: the travel clerk cannot be claiming a 'special relationship' with
every anonymous caller. Nor can she be claiming deep personal knowl-
edge of the addressee, associated, as I have argued, with 'closeness'.
What is signalled by her self-presentation, then, is neither 'intimacy' nor
'closeness', but rather characteristic Australian 'informality' - the same
informality which, for example, Australian university students express
by addressing their lecturers by their first name, or which Australian
public servants express by addressing their colleagues, and most of
their superiors, by their first names.
What is the meaning of this near-universal Australian 'informality'?
I think the essence of 'informality' (at least as practised in Australia)
lies in the purposeful rejection of any overt show of respect, with impli-
cations of familiarity, friendliness, and equality. Thus, by saying Cathy
speaking the travel clerk is inviting the anonymous callers to treat her
as if they knew her well, to assume that she 'feels something good
112 Cross-cultural pragmatics and different cultural values
towards all callers, the present caller included', and that there is no need
to show overt respect towards her (for example, by calling her Miss, Mrs,
or Ms). A university lecturer or a branch head who invites his or her
students or subordinates to address him or her as Bob or Jane conveys
a similar attitude. Very roughly:
(a) you don't have to 'show overt respect for me'
(b) I want you to speak to me as people do when they think:
(c) we know one another well
(d) we feel something good towards one another
(e) we can speak to one another in the same way
Component (c) of this explication implies familiarity, component (d),
mutual 'good feelings', and component (e), egalitarianism. Component
(b) shows that the speaker doesn't really have to know the addressee, to
have personal good feelings towards the addressee, or to claim full equal-
ity and full symmetry in his or her relation with the addressee (for
example, the travel clerk may well call the addressee Mrs Brown or Dr
Smith while calling herself Cathy). By using one's first name, or the
addressee's first name, the speaker is evoking a certain prototype of
human relations (spelled out in the components (c), (d), and (e)), and
this is, I suggest, the essence of 'informality'. In addition, however,
'informality' has to be opposed to 'formality'; this is reflected in the
following, additional component:
(f) I know: people can't always speak like this to other people
'Formality' is not always associated with hierarchical human relations
and with anti-egalitarianism. For example, in Australia, at formal meet-
ings of a university faculty, everybody speaks in a very 'formal' way,
without dissociating themselves thereby from the Australian ethos of
super-egalitarianism.
In Polish culture, titles of respect are used widely, and mutually:
'informality' is not valued in the way it is in Australia; yet this relative
'formality' is linked with a democratic, relatively egalitarian ethos (cf.
Davies 1984:331-336). On the other hand, in 'vertical' societies such as
Korea or Japan (cf. Nakane 1972), the value placed on social hierarchy
is closely linked with value placed on 'formality'. Hence, from a Korean
or Japanese perspective, the 'informality' of the Australian or American
culture may seem to be linked to their egalitarianism even more closely
than it really is.
Further illustrations: same labels, different values 113
In fact, 'informality' does tend to be linked with egalitarianism, and
'hierarchy' does tend to be linked with 'formality'; but none of these
links is straightforward, and none of them can be understood outside the
whole complex of other cultural norms and values of a given society.
Above all, the norms themselves have to be well understood and
carefully defined.
3.4. 'Harmony'
In the discussion of 'closeness' I have used repeatedly the word 'har-
mony' - another key word used widely in discussions of cross-cultural
pragmatics. But of course 'harmony' is no more self-explanatory than
,self-assertion', 'indirection', 'intimacy' or 'closeness', and if one
doesn't say what one means by it in a particular context, it can be as
misleading as the other widely used global labels. In the literature, this
word, too, has been used in many different and mutually incompatible
senses. For example, while both Anglo-American and Japanese cultures
can be said, and have been said, to value 'harmony', it is clear the
Anglo-American culture doesn't aim at 'harmony' in the sense in which
Japanese culture does; in particular, it doesn't aim at sameness, or
apparent sameness, of thoughts. Patricia Clancy describes this Japanese
view of harmony as follows:
The Japanese reliance upon indirection is consistent with their attitude
towards verbal conflict. As Barnlund points out, in Japan conversation is
'a way of creating and reinforcing the emotional ties that bind people
together' with the aim of social harmony. Therefore, overt expression of
conflicting opinions is taboo. Even conference participants ... , in contrast
to their argumentative American counterparts, tend to express their views
tentatively, in anticipation of possible retraction or qualification depending
upon how they are received; they try to feel out the positions of their
colleagues, seeking a common ground for establishing unanimity
(Barnlund 1975[b]; Doi 1974).
. .. Individuals may hold their own view, but, in the interests of group
harmony, should not express it if it conflicts with the opinion of others.
(Clancy 1986:215)
The following comment offered by Clancy (with reference to Doi 1974)
sums up the Japanese approach to harmony in a particularly striking
way:
114 Cross-cultural pragmatics and different cultural values
Since Japanese is a left-branching verb-final language, with negation
appearing as a verb suffix, speakers may negate a sentence at the last
moment, depending upon the addressee's expression. (Clancy 1986:214)
This attitude, which is certainly different from the Anglo-American
idea of harmony, can be portrayed as follows:
when someone says something
I can't say: 'I don't think the same'
someone could feel something bad because of this
when people say: 'we all think the same' it is good
As Clancy mentions above, Barnlund (1975b) links the attitude
portrayed here with the aim of 'creating and reinforcing the emotional
ties that bind people together'. This sounds rather like the Polish or
Russian ideal of 'closeness'. In fact, however, the attitudes involved are
almost diametrically opposed. In Slavic culture, saying 'I don't think the
same' is seen as promoting rather than jeopardising 'closeness'; and
'causing people to feel something bad' (now) can be seen as promoting
'closeness' in the long run.
Anglo-American attitudes to 'harmony' and 'closeness' are different
again. When considered from a Slavic or East European point of view,
Anglo-American culture must be said to be oriented to 'harmony' rather
than 'closeness', but certainly not to the kind of 'harmony' sought in
Japanese culture. Obviously, Anglo-American culture does not discour-
age people from saying 'I don't think the same'. It does, however,
discourage them from saying 'what you think is bad', 'I don't want you
to think this', 'I think something bad about you', and so on. Furthermore,
it doesn't encourage people to say things which are likely to cause the
addressee to 'feel something bad' (not even temporarily, in the interests
of long-term 'closeness').
Anglo-American attitudes to unanimity and 'harmony' were spelled
out earlier:
I can say what I think
you can say what you think
we don't have to think the same
this is good
(no one has to feel something bad because of this)
Polish attitudes to 'harmony' and 'closeness' are epitomised in the
proverb
Further illustrations: same labels, different values 115
Kto ~ czubi ten ~ [ubi.
'People who peck one another on the head (like fighting birds)
like one another.'
As this proverb suggests, it is not only difference of opinions which
is valued in Polish culture, but a forcefully, pointedly, and painfully
expressed difference. This attitude can be portrayed as follows:
I want to say what I think
I know: you can feel something bad because of this
I don't want not to say it because of this
I want you to know what I think
Once again, we must conclude that global labels such as 'harmomy'
or 'distance' obscure rather than clarify the real differences between
different cultures and different ethnographies of speaking.
3.5. 'Sincerity'
The problem of 'closeness' in interpersonal relations is closely related to
the problem of sincerity. It has often been said that in modem Western
culture sincerity has emerged as one of the core values. For example,
Trilling writes:
If sincerity is the avoidance of being false to any man through being true to
one's own self, we can see that this state of personal existence is not to be
attained without the most arduous effort. And yet at a certain point in
history certain men and classes of men conceived that the making of
this effort was of supreme importance in the moral life, and the value
they attached to the enterprise of sincerity became a salient, perhaps a
definitive, characteristic of Western culture for some four hundred years.
(Trilling 1972:5-6)
This may be so - but what does this crucial norm of 'sincerity' really
mean? Trilling (1972:2) offers the following definition: "The word as we
now use it refers primarily to a congruence between avowal and actual
feeling". We could translate this definition into the following formula:
if I don't feel X I shouldn't say 'I feel X'
Is it true that the norm spelled out above is an important feature of
Western culture? More specifically, is it true that it is an important
feature of Anglo-American culture?
116 Cross-cultural pragmatics and different cultural values
It is interesting to note that the subjective experience of Eastern-
European immigrants in English-speaking countries often leads them
to the opposite conclusion. In particular, Eastern European immigrants
often complain about what they perceive as the 'insincerity' of English
conversational routines, above all, of conversational openings such
as How are you?, Nice to see you, Lovely day, isn't it, and so on. (Cf.
Drazdauskiene 1981).
The perceived 'insincerity' of the 'How are you?' routine consists
both in the belief that the speaker doesn't really want to know how the
addressee feels and is expecting the addressee to reply positively ('Fine,
thank you', 'Very well, thank you', 'Not too bad') regardless of how the
addressee really feels. Consequently, common positive answers ('Fine,
thank you') are felt to be generally insincere, and the whole game is
perceived as an exercise in shared insincerity. I can add to this my
personal testimony as an immigrant and a bilingual: After seventeen
years of living in Australia, I still find the pseudo-question How are you?
a perplexing one, since my own cultural impulse is to try to reply
sincerely, which I know I am not supposed to do. When I recently failed
to reply promptly to this question, helplessly searching for words, my
interlocutor laughed at me: 'Come on, this is not such a difficult
question'. But to me, it is a difficult question, and I know that I share
this difficulty with thousands of other East European immigrants in
Australia and in America.
I can't believe, therefore, that Anglo-American culture really cher-
ishes and promotes the norm that Trilling attributes to it: 'if I don't
feel X I shouldn't say "I feel X"'. On the contrary, I think that Slavic
and Eastern European culture promotes this norm, and that by doing so
it comes into conflict with Anglo-American culture.
But this is not to say that I don't recognise the validity of what
Trilling is trying to say about Western culture (as opposed to what he
actually does say). Clearly, what he had in mind was not sincerity (or
otherwise) of conversational formulae, but sincerity of certain kinds of
self-disclosure. Trilling (1972:5) quotes in this connection Matthew
Arnold's "wistful statement of the difficulty, perhaps even impossiblity,
of locating the own self':
Below the surface-stream, shallow and light,
Of what we say we feel - below the stream,
As light, of what we think we feel - there flows
With noiseless current strong, obscure and deep,
The central stream of what we feel indeed.
Further illustrations: same labels, different values 117
Matthew Arnold called the hidden self the 'best self', but, Trilling
asks, 'is it the own self?' In Trilling's (1972:5) view, if there is anything
deep down in me which corresponds to 'the archetype of human being',
that is, to the 'mankind's best self', this is not my sole self: "I know
that it coexists with another self which is less good in the public moral
way but which, by very reason of its culpability, might be regarded as
more peculiarly mine. So Hawthorne thought: 'Be true! Be true! Be true!
Show freely to the world, if not your worst, yet some trait by which the
worst may be inferred.' "
This brings us, I think, much closer to the real meaning of 'sincerity'
in European culture. It is not a question of never saying that one feels
something that one doesn't feel; rather, it is a question of knowing what
one really feels (including feelings that reveal something bad about
oneself) and of being able to disclose those real feelings (especially
those which show something bad about oneself) 'to the world'. Every
human being is unique, and uniquely interesting because of this. We
shouldn't try to appear 'good' to other people. Rather, we should try to
reveal 'to the world' our uniqueness, and this involves, above all else,
our 'badness': because our 'badness' is more original, and more interest-
ing, than our 'goodness'.
The cultural injunctions in question can be formulated as follows:
I don't know what I feel
I want to know it
when I know it I want to say it
I want people to know it
I think that people can think something bad about me because
of this
I don't want not to say it because of this
I believe that the attitude spelt out above may indeed be, as Trilling says,
'a salient, perhaps a definitive characteristic of Western culture', linked
closely with the birth of Western individualism, with the emergence and
growing significance of mirrors, self-portraits, diaries, autobiographies,
'confessions', introspection, and so on. My point is that we cannot
capture, or identify, this characteristic by means of some global term
such as 'sincerity'.
Contemporary Anglo-American culture doesn't seem to place any
premium on 'never saying that one feels something that one doesn't
feel'. On the contrary: the routines of human interaction reflected in the
English language encourage saying that one feels something good when
118 Cross-cultural pragmatics and different cultural values
one doesn't feel anything good. This is manifested, in a spectacular
way, not only in the How are you? routine, but also in the conventions
of letter-writing: the opening phrase Dear Sir expresses a good feeling
towards an addressee who may be a complete stranger, and so does the
closing phrase Yours sincerely. Phrases of this kind cannot be used in
other European languages, certainly not in Slavic languages, which do
not allow any formalised expression of clearly non-existent 'good feel-
ings' (and which encourage expression of existing 'bad feelings ').
Trilling (1972:3) opens his discussion of 'sincerity' in Western culture
with a quote from Hamlet:
This above all: to thine own self be true
And it doth follow, as the night the day,
Thou canst not then be false to any man.
It would appear, however, that in modem times the idea of 'being true
to oneself' has become disassociated from that of 'never being false
to any man'. This may be linked with the shift of emphasis from 'sincer-
ity' to 'authenticity', which as Trilling points out, has taken place in
modem times.
A very considerable originative power had once been claimed for sincerity,
but nothing to match the marvellous generative force that our modem
judgment assigns to authenticity ... Still, before authenticity had come
along to suggest the deficiencies of sincerity and to usurp its place in our
esteem, sincerity stood high in the cultural firmament and had dominion
over men's imagination of how they ought to be. (Trilling 1972: 12).
It seems to me that in modern times the two ideas linked by Shakespeare
in the passage from Hamlet have become dissociated: the idea of 'being
true to oneself' developed into something like that 'authenticity' dis-
cussed by Trilling, whereas the idea of 'not being false to any man'
has given way to a modern Anglo-American virtue of social harmony,
based on 'distance' and on avoidance of interpersonal clashes.
The virtue of 'authenticity' has to do with the notion of 'self', and of
a true, genuine, and uninhibited expression of one's self. It does not in-
volve the relation between 'I' and 'you'. As far as the relation between
'I' and 'you' is concerned the emphasis seems to have shifted from
'sincerity' to the avoidance of clashes, to smooth, well-greased, harmo-
nious social interaction. Conventional expressions and conventional
routines such as Dear Mr X, How are you?, Lovely to see you, Nice to
have met you, Lovely day, isn't it, and so on, provide the oil for such
harmonious social interaction. The expansion of such expressions fits
Further illustrations: same labels, different values 119
in logically with the modern Anglo-American constraints on direct
confrontation, direct clashes, direct criticisms, direct 'personal remarks'
- features which are allowed and promoted in other cultures, for
example, in Jewish culture (cf. Schiffrin 1984) or in Black American
culture (cf. for example Kochman 1981), in the interest of cultural values
such as 'closeness', 'spontaneity', 'animation', or 'emotional intensity',
which are given in these cultures priority over 'social harmony'.
This is why, for example, one doesn't say freely in (white) English,
'You are wrong', as one does in Hebrew (cf. Schiffrin 1984) or 'You're
crazy', as one does in Black English (cf. Kochman 1981:46). Of course
some 'Anglos' do say fairly freely things like Rubbish! or even Bullshit!.
In particular, Bullshit! (as well as You bastard!) is widely used in con-
versational Australian English. Phrases of this kind, however, derive
their force and their popularity partly from the sense that one is violating
a social constraint. In using phrases of this kind, the speaker defies a
social constraint, and exploits it for an expressive purpose; indirectly,
therefore, he (sometimes, she) acknowledges the existence of this
constraint in the society at large.
Generally speaking, in mainstream Anglo-American culture one has to
be rather careful as to what one says about 'you' (because one prefers
to avoid confrontation, preserve harmony, avoid the impression of
imposing or interfering, and so on); at the same time, one can be less
circumspect in saying things about oneself, although here, too, there are
various constraints and restrictions, such as the constraint on 'bragging'
or the constraint on the expression of one's bad feelings towards the
addressee, or the constraint on 'emotional displays'.
The general norm, then, can be portrayed as follows:
I can say what I think/want/feel
other people can say what they think/want/feel
Some of the constraints on this general norm can be formulated as
follows:
(1) I can't say: I am good
I can do things that other people can't
(people would think something bad about me
because of this)
(2) I can't say: I feel something bad towards you
I think something bad about you
120 Cross-cultural pragmatics and different cultural values
(3) I can't say: I want you to do something that you don't want
to do
(4) I can't always say what I feel
(people could think something bad about me because of this)
(5) I can't say things when someone else is saying something
One might say, then, that a curious paradox is involved in the position
of 'sincerity' as a cultural value in modem Anglo-American culture. On
the one hand, as Trilling says, the very word sincererly) has come
to have an air of insincerity about it - and, yet, as Goldstein - Tamura
(1975) point out, 'Anglos' (in contrast to the Japanese) go to great
trouble to sound sincere. To achieve this, they seek to express their
feelings in a personalised way, in contrast to the Japanese, who rely on
standard forms and do not view 'cliches' or ready-made formulae in a
negative way:
To the American, the Japanese method of standard messages, such as
'Congratulations' with only a name, the presentation of a gift with a stan-
dard phrase, a refusal with a standard phrase before acceptance ... may
seem very bare indeed and perhaps somewhat insincere. (Goldstein -
Tamura 1975:91) [emphasis added]
The American guest expressing thanks to his host at the end of dinner has
no ... standard form, but rather makes use of a variety of possibilities
generally emphasising the success of the meal, with or without an
expression of thanks, such as 'Thank you for the delicious dinner' or
'What a delicious meal that was' (more informal) or 'Boy, that was great!'
(colloquial-slang), each said with appropriate intonation to express sincer-
ity. (Goldstein - Tamura 1975:72) [emphasis added]
This search for a personalised expression of, so to speak, predictable
feelings, seems to reflect a tension between the value of authenticity,
of 'being true to oneself', and the search for friendly, harmonious
relations with other people; between the desire to express one's 'real
self' ('this is what I feel/want/think') and the desire to have friendly
interpersonal relations with other people and to ensure that 'everyone
feels something good'.
The desire to have friendly relations with other people may lead one to
say things which do not correspond to what one really feels and thinks.
The awareness of this, and the value placed on both 'harmony' and self-
expression, may lead to an attitude which can be portrayed as follows:
Different attitudes to emotions 121
(a) I say: I feel something
(b) I say this because I feel this
(c) I know: you can think that I say this
because I think I should say it
(d) I don't want you to think this
(e) I say this because I feel this
In Japanese culture, there is no room for this kind of attitude because the
emphasis is very largely on saying what one thinks one should say, not
on saying what one really feels. Hence, there is no perceived need to use
a personalised form 'to ring true'.
. .. the American speaker makes a personal connection to the hearer [sic]
while at the same time expressing his own personality in the arrangement
of words he chooses to use. The Japanese speaker, having the form at his
disposal, shows chiefly his awareness of his obligation by using the verbal
form at the appropriate level at the appropriate time. (Goldstein - Tamura
1975:80)
4. Different attitudes to emotions
Different cultures take different attitudes to emotions and these different
attitudes to emotions influence, to a considerable degree, the ways
people speak. (See for example Lutz 1986, 1988; Wierzbicka, to appear.)
Differences of this kind cannot be satisfactorily explained by means of
any global labels such as 'emotional' or 'anti-emotional', 'expressive' or
'non-expressive'. They can, however, be made clear by means of
semantic explications. In what follows, I will try to present thumb-nail
sketches of several cultures, considered from the point of view of their
characteristic attitudes to emotions.
4.1. Polish culture
Like other Slavic cultures, Polish culture values what might be called
uninhibited emotional expression:
I want to say what I feel
122 Cross-cultural pragmatics and different cultural values
This includes both good feelings and bad feelings:
I feel something good/bad
I want to say it
As mentioned earlier, it values in particular expression of good feelings
towards the addressee:
I feel something good towards you
(I want you to know it)
and it offers for this purpose a wealth of linguistic resources, such as
a very rich system of hypocoristic forms of personal names, and also a
rich set of terms of endearment. The latter point is illustrated by Polish
terms widely used in everyday speech, particularly in speech directed
at children: ptaszku 'dear little bird', kotku 'dear little cat', sloneczko
'dear little sun', iabko 'dear little frog', skarbie 'treasure', z/otko 'dear
little gold', and so on (cf. Wierzbicka, to appear).
Many other features of the Polish ethnography of speaking can be
explained in terms of this cultural attitude. For example, cordial impera-
tives and 'impositives' ('have some more', 'you must have some more',
'you must stay a little longer', and so on) are clearly related to it
(cf. Chapter 2 above).
Polish principle of 'cordiality'
I feel something good towards you
I want good things to happen to you
I want to be with you
4.2. Jewish culture
Emotional self-expression was also highly valued in traditional (East
European) Jewish culture, as described, for example, by Matisoff (1979).
In this culture, however, good and bad feelings were generally expressed
by means of good and bad wishes. Hence the tremendous importance
of curses and blessings in Yiddish speech. This characteristically
Jewish style of emotional expressiveness is well illustrated in the
following passage:
There are as many types of curses as there are people cursing, but the
hardest to explain is the mother cursing her child. The child may be crying
Different attitudes to emotions 123
because he is hungry. The mother bursts out, 'Eat, eat, eat. All you want to
do is eat. May the worms eat you. May the earth open up and swallow you
alive.' This mother loves her child, she is only pouring out the bitterness
that's in her heart in the only way she knows. But in translation she sounds
like a monster. (Butwin 1958:9)
A few further examples of Jewish 'wishes' expressing the speaker's
feeling (see Matisoff 1979):
Governor Reagan, may he be erased, isn't giving any raise this
year to my son the professor, a health to him.
A black year on her, all day long she chewed my ear off with
trivia.
My mother-in-law, maya lament be known to her, has a wicked
tongue.
My wife - must she live? - gave it away to him for nothing.
Matisoff (1979:86) offers the following comment, which I believe
expresses a deep insight: "Especially in the case of curses, the formulas
may serve a purely therapeutic function. They are convenient, con-
ventionalised ways of letting off steam - releasing bursts of psychic
energy which might otherwise remain hopelessly bottled up ... " Follow-
ing this insight, we can represent the pragmatic principle in question
as follows (the bracketed component is optional):
Jewish expressive curses
X thinks of person Y
(X thinks: person Y did something bad)
X feels something because of this
X wants to say something because of this
X says: I want something bad to happen to Y
4.3. American black culture
Uninhibited emotional self-expression is also characteristic of American
black culture, as opposed to white culture. In this culture, however, there
is no emphasis on 'good feelings towards the addressee', and there is no
tradition of expressive wishes. Among several characteristic features
of this culture which emerge from the rich literature on the subject I
will single out the 'intense', 'emotional' character of black speech: the
124 Cross-cultural pragmatics and different cultural values
'animation', the 'heated tone' of discussions, the 'lack of detachment' in
stating one's opinions and expressing one's thoughts, even on abstract,
intellectual topics; the distrust of a deliberately dispassionate and
detached mode of discussion favoured by white Anglo-American
culture. Kochman writes:
But they [blacks] have another reason to misinterpret (and distrust) the
dispassionate and detached mode that whites use to engage in debate. It
resembles the mode that blacks themselves use when they are fronting: that
is, consciously suppressing what they truly feel or believe. As one black
student put it, 'That's when I'm lyin'.' Fronting generally occurs in black/
white encounters when blacks perceive a risk factor and they decide it
would be more prudent to keep silent than to speak. (Kochman 1981:22)
Black culture values, then, and promotes the following attitude:
I think something
I feel something because of this
I want to say it
White Anglo-American culture values and promotes what one might
call loosely the opposite attitude:
I think something
I want to say it
I don't feel anything because of it
The two cultural norms in question could also be represented as follows:
Black American
X thinks something
X wants to say it
one can see that X feels something because of this
people think: this is good
Anglo-American
X thinks something
X wants to say it
one can't see that X feels anything because of this
people think: this is good
Furthermore, for blacks, views are inseparable from values and values
are closely linked with emotional involvement. Consequently, "blacks
present their views as advocates. They take a position and show that
they care about this position" (Kochman 1981 :20) - and they care about
Different attitudes to emotions 125
it because they think it is good. By contrast, whites tend to present their
ideas as spokesmen, not advocates. "How deeply a person cares about
or believes in the idea is considered irrelevant to its fundamental value.
. .. Whites believe that caring about one's own ideas, like the infatuation
of scientists with their own hypothesis, will make them less receptive to
opposing ideas." (1981:21). Kochman speaks in this connection of the
separation of 'truth' and 'belief' in Anglo-American culture, and he links
the norms of dispassionate, neutral objectivity, and of detachment from
one's ideas with the desire to discover 'the real truth' (which is seen as
involving only sentences, not people). He points out that in this culture
the merits of an idea are seen as intrinsic to the idea itself, and that
emotional involvement with ideas is seen as something that can only
prevent people from being able to assess their intrinsic values.
Black American
I think something
I feel something because of this
I think it is good to think this
I want other people to think this
Anglo-American
I think something
I don't feel anything because of this
I know other people don't have to think the same
I want to say what I think
I want other people to think about it
I want to know what other people think about it
The conviction that our ideas are good and the attitude of emotional
attachment to them leads in black culture to what Kochman (1981:23)
calls 'dynamic opposition', an attitude which is perceived as a unifying
rather than a divisive force. "Whites attempt to minimise dynamic
opposition within the persuasive process because such confrontation, or
struggle, is seen as divisive. Blacks, however, see such struggle as
unifying ... It signifies caring about something enough to want to
struggle for it." I think that to account for this 'dynamic opposition' and
for its 'unifying force' we could add to the formulae sketched above the
following components:
Black American
I know that you don't think the same
126 Cross-cultural pragmatics and different cultural values
I think this is bad
I feel something because of this
I want us to think the same
Anglo-American
I know that you don't think the same
I don't think this is bad
I don't feel anything because of this
I think: we don't have to think the same
4.4. Japanese culture
In Japanese culture, as we have seen, the prevailing norm with respect
to emotions is this:
I don't want someone to feel something bad
This is manifested in countless ways in the Japanese ethnography of
speaking, but perhaps most spectacularly in the omnipresence of
apologies and quasi-apologies in Japanese speech (cf. Coulmas 1981;
cf. also Mizutani - Mizutani 1987). The importance of apologies in
Japanese culture is epitomised in the fact that in the Japanese version
of Little Red Riding-Hood the wolf has to appear at the end with tears
in his eyes asking for forgiveness (Lanham 1986:290). The theme of
indebtedness which pervades Japanese social interaction is related to this
omnipresence of apologies, and also to the lack of boundaries between
acts which from a Western perspective would be interpreted as apologies
and thanks. Roughly:
(1) I did something (that was bad for you)
I think you could feel something bad because of this
I feel something bad because of this
(2) You did something good for me
I didn't do something like this for you
I feel something bad because of this
Thus, both the constant fear that someone may feel something bad
because of us and the constant awareness of unrepaid good things
that other people have done to us lead to the humble expression of our
own 'guilt':
I did/didn't do something (to/for you)
Different attitudes to emotions 127
I feel something bad because of this
In addition, Japanese culture places very high value on empathy, on
anticipating what other people might feel. It is this high sensitivity to
other people's (unexpressed) feelings that causes the Japanese to 'mask'
and conceal their feelings:
In social interaction, Japanese people generally are expected to restrain, if
not suppress, the strong or direct expression of emotion. Those who cannot
control their emotion are considered to be immature as human beings.
Strong expression (verbal or nonverbal) of such negative emotions as an-
ger, disgust, or contempt could embarrass other people. Direct expression
of sorrow or fear could cause feelings of insecurity in other people.
Expression of even happiness should be controlled so that it does not
displease other people.
The best way to comply with this social code of behavior is to utilise
masking techniques. Thus, Japanese people, although unaware, frequently
display apparent lack of a meaningful facial expression, often referred to
as 'inscrutable' by Western people. It is an attempt to neutralise strong
emotions to avoid displeasure or embarrassment on the part of other
people. (Hanna - Hoffer 1989:88-90)
This can be represented as follows:
I don't want to say what I feel
someone could feel something bad because of this
This focus on empathy is manifested not only in the constant attempts
to avoid anything that might hurt or offend the addressee, but also in the
attempts to anticipate and guess other people's unexpressed needs and
wishes (cf. Lebra 1976).
The Japanese principle of 'empathy' and 'consideration'
X thinks:
if I do something (Y)
this person can feel something bad/good because of this
I will not/will do it because of this
this person doesn't have to say anything
This stress on empathy is linked to the Japanese reluctance to verbal-
ise feelings at all - due largely to the fear that by doing so one may
hurt or offend other people, but also to the conviction that feelings
should be expressed, and understood, without words; and furthermore,
that feelings cannot be really expressed by words. Goldstein and Tamura
128 Cross-cultural pragmatics and different cultural values
comment in this connection:
In general, ... Americans ... tend more to the feeling that they can express
sentiments directly and personally through the medium of language.... To
the Japanese, on the other hand, 'No word conveys sorrow; you must look
at the color of the eyes'. (Goldstein - Tamura 1976:92-93)
The attitude epitomised in this proverb can be portrayed, roughly, as
follows:
I don't want to say what I feel
one can't say what one feels
4.5. Javanese culture
As we have seen earlier, not saying what one feels is also highly valued
in Javanese culture. Here, however, the motivations seem to be rather
different than in Japan: it is not so much the belief that feelings cannot
be expressed in words, or the preference for wordless empathy, or the
consideration for other people's feelings, but rather, a desire to protect
one's own equanimity and peace of mind, which could be threatened by
an overt expression of feeling. Thus, Geertz writes:
If one can calm one's most inward feelings (by being trima, sabar, and
iklas), ... one can build a wall around them; one will be able both to
conceal them from others and to protect them from outside disturbance.
The refinement of inner feeling has thus two aspects: the direct internal
attempt to control one's emotions represented by trima, sabar, and iklas;
and, secondly, an external attempt to build a wall around them that will
protect them. On the one hand, one engages in an inward discipline, and
on the other in an outward defence. (Geertz 1976:241)
This can be reflected in the formula:
I don't want people to know what I feel
It would seem, then, that while neither the Japanese nor the Javanese
want to say what they feel, the Javanese, in addition, don't even want
others to know what they feel; and they want to restrain not only the
external expression of feelings, but also internal emotional experience
(an attitude reminiscent of the Stoic apatheia 'freedom from emotional
disturbance', see Wierzbicka, to appear, chap. 6).
Conclusion 129
The management of one's emotional economy becomes one's primary
concern, in terms of which all else is ultimately rationalised. The spiritu-
ally enlightened man guards his psychological equilibrium well and makes
a constant effort to maintain its placid stability. His proximate aim is
emotional quiescence, for passion is kasar feeling, fit only for children,
animals, peasants, and foreigners. His ultimate aim, which this quiescence
makes possible, is gnosis, the direct comprehension of the ultimate rasa.
To feel all is to understand all. Paradoxically, it is also to feel nothing ...
Emotional equanimity, a certain flatness of affect, is, then, the prized
psychological state, the mark of the truly alus [refined] character.
(Geertz 1976:239-240)
This would suggest an attitude which can be portrayed as follows:
I want to feel the same all the time
I think if I do something this will happen
I think I can do this
5. Conclusion
For intercultural understanding, "More than mere contact is essential.
People must become capable of empathy, of being able to project them-
selves into the assumptive world, the cultural unconscious, of an alien
culture. Yet this is a formidable task unless there are ways to introduce
people to the assumptive world of others" (Bamlund 1975b:140).
I have tried to show that there are ways to do this. We cannot enter
the 'assumptive world of others' if we try to rely on culture-specific,
complex, and obscure concepts such as 'directness', 'self-assertion',
'solidarity' or 'harmony'; but one can do it if we rely, instead, on lexical
universals such as want, think, say, or know.
Ruth Benedict (quoted in Barnlund 1975b:140) wrote:
One of the handicaps of the twentieth century is that we still have the
vaguest and most biased notions, not only of what makes Japan a nation of
Japanese but of what makes the United States a nation of Americans,
France a nation of Frenchmen, and Russia a nation of Russians. Lacking
this knowledge, each country misunderstands the other. (Benedict 1947: 13)
What applies to different nations, applies also to different ethnic groups
in a multiethnic society. What makes Japan a nation of Japanese, or
130 Cross-cultural pragmatics and different cultural values
Russia a nation of Russians is reflected - more clearly than anywhere
else - in the ways the Japanese or the Russians speak. And the ways
in which they speak can be summarised in clearly and universally
accessible formulae couched in the natural semantic metalanguage.
Chapter 4
Describing conversational routines
It is a truism to say that different cultures, and subcultures, have differ-
ent conversational routines, and that it is important that those different
routines should be carefully studied, analysed and described. But the
ways in which this self-evident program of research should be imple-
mented are by no means clear or generally agreed upon. In this chapter I
shall argue that despite the considerable effort which has gone into the
description of conversational routines, much less has been achieved in
this important area than might have been - because not enough thought
has been given to the vital question of a metalanguage in which such
analysis can be fruitfully carried out.
To show how a suitable metalanguage can facilitate the description
and comparison of conversational routines, I examine a number of
generalisations suggested, or hinted at, in Anita Pomerantz's (1978)
interesting paper on responses to compliments. I try to show why in
the present form these generalisations are neither clear nor verifiable,
and I propose ways of reformulating them which could make them clear
and verifiable. I try to show how the use of the proposed metalanguage
makes such reformulation possible, and how it enables us to describe
conversational routines used in different societies in a way which can be
illuminating, rigorous and free of ethnocentric bias. (For another attempt
along similar lines, see Ameka 1987.)
1. Conversational analysis: linguistic or non-linguistic
pragmatics?
Many conversational routines in many cultures are lexicalised and/or
grammaticalised, that is to say, they consist in uttering in certain
situations certain phrases, or using certain constructions, which encode
certain language-specific interactional meanings. It seems clear that
132 Describing conversational routines
meanings of this kind have to be revealed and described - like any other
kind of meanings. This is the task of linguistic pragmatics.
For example, English conversations are often started with the con-
ventional phrase How are you? Leech (1983:198) comments on the
meaning of this phrase by quoting a couplet:
Don't tell your friends about your indigestion:
'How are you!' is a greeting, not a question.
But as this couplet implicitly recognises, How are you? is not just a
greeting, but a kind of cross between a greeting, a question, and an
invitation for the addressee to say something about their current state -
something that is expected to be short and 'good' rather than long
and 'bad'.
But this kind of description, though useful as a starting point, is
very imprecise and, what is worse, it is inherently ethnocentric. English
words such as greeting, question and invitation belong to the English
folk-taxonomy of speech acts and have no exact equivalents in other
languages, so they cannot possibly be regarded as useful analytical tools
for cross-cultural comparison. On the other hand, useful tools of this
kind can be found in relatively simple words such as say, want, know,
someone, something, good or bad, which have their semantic equivalents
in all (or nearly all) languages of the world. Using such simple and (rela-
tively) culture-free tools, we can formulate the meaning of the English
phrase How are you? along the following lines:
How are you?
(a) I know: we can now say things to one another
(because we have come to be in the same place)
(b) I want to say something to you because of this
of the kind that people say to one another
when they come to be in the same place
(c) I want you to know: I feel something good towards you
(d) I say: I want to know 'how you are now'
(e) I want you to say something because of this
(f) 1 want you to say: 'I am well'
(g) 1 think you will say something like this
(h) I think we will feel something good because of this
Component (a) of this formula shows that the phrase in question is
a conversation opener, or a potential conversation opener. Component
(b) shows that the phrase constitutes an established linguistic routine
Conversational analysis: linguistic or non-linguistic pragmatics? 133
employed in such circumstances. Component (c) shows the friendly
character of the phrase (absent, for example, from the otherwise compa-
rable phrase Good-bye, cf. Wierzbicka 1987:224). Component (d) shows
the speaker's (real or pretended) interest in the addressee's well-being.
Component (e) indicates that How are you? - like a question - obliges
the addressee to make a verbal response. Component (f) indicates what
kind of response is expected (a positive one); and it may be taken to
indicate the speaker's wish that the addressee should be well (I want you
to say that you are well because 1 want to know that you are well,
because 1 want you to be well). Component (g) shows the speaker's
optimistic expectation that the answer will be positive (and betrays at
the same time a reluctance to hear a negative one). Component (h)
suggests that a positive answer will be a 'pleasure' to both interlocutors
and hints that this shared pleasure will be conducive to social harmony
between them.
A response to How are you? involves typically three steps (the first
being obligatory and the other two optional). First, one answers the
interrogative element of the utterance by saying something like 'I am
well' or 'I want to say: 1 am well (but 1 can't)'; second, one thanks the
interlocutor (or rather, one says thank you, or thanks); and third, one
reciprocates the act (And how are you?, And yourself?, etc.) Of these
three steps, the second seems largely lexicalised, that is, determined not
only semantically but also lexically. One cannot replace the expression
thank you with, for example, the expression thanks a lot:
A: How are you?
B: Very well, thanks *a lot.
Step three is determined in (some aspects of) its meaning, not in its form:
B: And you? / And yourself? / How about yourself?
Step one, too, is lexically 'free'. Semantically, the preferred response
is some phrase which means, essentially, 'I am very well'. Using for a
moment the kind of metalanguage employed by Pomerantz, we might
say that the addressee can be expected to answer in 'strongly positive
terms' or, if this is felt to be impossible, at least to avoid answering in
'strongly negative terms'. A sincere and spontaneous positive self-report
('I am well') will tend to be 'upgraded' to something like I'm very well
or I'm fine, and a sincere and spontaneous negative self-report ('I am not
well') will tend to be 'toned down' (or is it 'up'?) to something like Not
very well, I'm afraid.
134 Describing conversational routines
Emphatically negative answers like Rotten or Lousy! are possible, but
they are felt to violate the normal routine, and they tend to be delivered
in a jocularly defiant manner, implying: 'I know this is not what I am
expected to say'.
It must be acknowledged, however, that while the phrase How are
you? (as a conventional conversational opening) is part of the English
language and should be included in an adequate dictionary of English
(along with Good morning, Hello! and Hi!), the responses to this phrase
are not similarly conventionalised, and should not be similarly listed
in a dictionary. The range of such responses, and the normal strategies
for formulating them, should of course be described, as an important
part of the communicative competence of English speakers, but they
would lie outside the boundary between linguistic pragmatics and non-
linguistic pragmatics.
What I suggest, nevertheless, is that in a sense, proper methods of
semantic description are equally relevant to both types of pragmatics.
In particular, both types need a justified and 'culture-free' semantic
metalanguage. It is not only the meaning of set phrases such as How are
you? which has to be stated in a meaningful and well-justified semantic
metalanguage, but also the meaning of loose descriptive expressions,
such as 'an upgrading strategy', which despite their technical ring are
metaphorical and have no clear meaning, and are therefore not empiri-
cally verifiable. As a starting point for a rigorous description and for
possible verification, I propose the following generalisation. A reply to a
How are you? has to take into account the following set of assumptions:
(a) I know that you want me to say something good
(b) I know that you don't want me to say something bad
(c) I think that you think I will say something very good
(d) I think that you think I will not say something very bad
The first speaker's presumed wishes and expectations can of course be
violated in the addressee's response, but they cannot be ignored: the
addressee must realise that any such violation will be seen as a depar-
ture from the norm, and will be interpreted accordingly. The range of
expected responses generated by these guidelines can be described
as follows:
(A) I want to say something very good
[Very well; Fine; etc.]
Conversational analysis: linguistic or non-linguistic pragmatics? 135
(B) I want to say something good
[Good; I'm well; etc.]
(C) I can't say something very good
I don't want to say something very bad
[Not too bad; I'm OK; etc.]
(D) I can't say something good
I don't want to say something very bad
[Not too good; Not very well; etc.]
In addition to these expected conventional responses there is also the
slightly humorous strategy mentioned earlier, which involves a conscious
violation of the convention:
(E) I don't want to say what you think I will say
I don't want to say what you want me to say
I want to say what I think
[Rotten; Lousy; Terrible; etc.]
One final strategy which should perhaps be recognised as a separate
possibility could be represented as follows:
(F) I don't want to say something good
I don't want to say something bad
I don't want you to think I say something that I don't think
I can say something good
[Not bad; etc.]
On the surface, this type of response may be difficult to distinguish from
type (C), but the intention behind each of them is different. Type (F)
implies that one is well but that one doesn't want to sound insincere
and gushing; it can be interpreted, therefore, as a kind of understatement.
Type (C), on the other hand, implies that one is not well, but that one
doesn't want to complain; it can be interpreted, therefore, as a kind of
overstatement. As pointed out by Sharon Henschke (p.c.), the potential
ambiguity can often be resolved by means of an interjection, and/or
intonation. The utterance:
Oh, not (too) bad.
will be interpreted as implying 'I am not (quite) well' (type C), whereas
a bright and cheerful Not bad, or Not too bad, (accompanied by a grin)
will be taken to imply 'I am well'.
136 Describing conversational routines
It appears that different social groups of speakers of English tend to
favour different response strategies. For example, I would hypothesise
that strategy (A) is used more frequently by women than by men; and
also, that it is used more frequently by Americans than by Australians;
and that the reverse would be true of strategies (E) and (F) (cf. Renwick
1980). These are of course matters for empirical investigation. But
they could not be empirically investigated if we didn't frame the initial
hypotheses in a rigorous and (at least relatively) culture-free manner.
2. 'Compliment response' routines
Turning now to Pomerantz's (1978) 'responses to compliments', we must
ask, first of all, what exactly she means by 'compliments', for it seems
fairly clear that she is not using this word in the ordinary sense. For
example, she includes among 'compliments' utterances such as the
following ones:
B: Well anyway nice talking to you
A: Nice talkin to you honey
(Pomerantz 1978: 107)
Furthermore, she often appears to be using interchangeably words
such as 'compliments', 'praise' or 'credit', which in the English folk-
taxonomy of kinds of speech acts stand of course for different categories
(cf. Wierzbicka 1987).
I presume that what Pomerantz really has in mind is a class of
speech acts which can be characterised in terms of the following
semantic component:
I want to say something good about you
She suggests that the addressee's problem consists in responding to such
an utterance in a polite and 'supportive' manner while at the same time
avoiding explicit or implicit 'self-praise'. We can reformulate this by
portraying the addressee's (expected) attitude as follows:
I don't want to say something good about myself
What can an addressee do to implement these conflicting guidelines
without being impolite or 'uncooperative'?
'Compliment response' routines 137
To answer this question, Pomerantz introduces a number of theoretical
constructs such as 'agreements' and 'disagreements', 'upgrades' and
'downgrades', and 'referent shifts'. These are introduced by means
of illustrations rather than any clear definitions. It appears, however,
that the essence of what is intended can be captured in the following
formulae:
'agreements'
A:X
B: I think the same
'disagreements'
A:X
B: I don't think the same
'upgrades'
A: X is good
B: X is very good
'downgrades'
A: X is very good
B: X is good
'referent shifts'
A: I want to say something (Y) about X
B: I want to say it (Y) about something other than X
One cannot be sure, of course, that these formulae do in fact correspond
to what Pomerantz had in mind, because those intentions have only
been hinted at in loose and metaphorical terms. They could be, however,
easily revised and amended if this proved necessary or desirable. The
important thing is that they are explicit, and that they force the analyst
to be explicit and to make clear analytical decisions. For example, a
term such as 'upgrade' may seem intuitively intelligible, but it doesn't
make it clear whether it is meant to apply only to a substitution of 'very
good' for 'good', or whether it is also meant to stand for a substitution
of 'very bad' for 'bad', or 'good' for 'not bad', and so on.
Trying in this way to tentatively translate Pomerantz's vague and
metaphorical constructs into a controlled semantic metalanguage, I now
proceed to consider some of her suggested generalisations in more detail.
138 Describing conversational routines
2.1. Upgrades
Pomerantz introduces the concept of 'upgrade' as follows:
One type of agreement, an upgrade, can be called 'optimal' on sequential
grounds. Upgrades are prevalent in environments in which agreements are
preferred; they occur in agreement terms and sequences and typically not
in combination with disagreements. Upgrading techniques include the in-
corporation of stronger second evaluation terms, for example ...
[1] A: Isn't he cute?
B: O::h he::s a::dorable
[2] A: She seems like a nice little lady.
B: [Awfully nice little person}
(Pomerantz 1978:93)
What exactly is an 'upgrade', then?
Pomerantz uses the expression 'stronger evaluation terms', but it
appears that in fact she means a 'good' evaluation rather than 'bad'. It
seems, therefore, that essentially, an 'upgrade' consists in a replacement
of 'good' by 'very good'. Since, however, in the quote adduced above
'upgrades' as a conversational strategy are linked with agreements (and
are even called a type of agreement) one is led to conclude that the
strategy which Pomerantz has in mind should in fact be represented
as follows:
A: I think X is good
B: I think the same
I would say X is very good
The component 'I think the same' spells out what Pomerantz calls 'agree-
ment', whereas the core of the 'upgrade' consists in replacing the first
speaker's 'good' with the combination 'very good'.
It is interesting to note in passing that the term 'upgrade' or 'upgrader'
is used by other analysts in what appears to be a totally different sense:
Request tokens were further analysed for the presence of linguistic
elements that serve to mitigate, soften, or 'downgrade' the act, or those
that serve the opposite function, that is, 'upgrading' aggravating elements.
(Blum-Kulka - Danet - Gherson 1985:119)
The authors regard this 'definition' as self-explanatory, and as suffi-
ciently precise to take it as a basis for counting: "The data showed
94 cases of downgraders and 118 cases of upgraders." (1985:119) This
'Compliment response' routines 139
shows, once again, how unreliable terms of this kind are, even when
they are backed with seemingly precise statistical data.
2.2. Contrastive opposites
Another conversational strategy singled out by Pomerantz is the 'con-
trastive opposite'. If an 'upgrade' constitutes 'an optimal agreement'
(presumably, in the context of evaluations), a contrastive opposite
constitutes 'an optimal disagreement'. Two examples:
(1) A: Did she get my card?
B: Yeah she gotcher card.
A: Did she t'ink it was terrible?
B: No she thought it was very adohrable.
(2) A: [ was wondering if [' d ruined yer - weekend by uh
B: [No. No. No, [ just loved to have - ...
The examples provided suggest that a 'contrastive opposite' can be
defined as follows:
A: I think my X is very bad
B: I don't think the same
I think your X is very good
Pomerantz (1978:93) writes: "Contrastive opposites are produced in
environments in which disagreements are preferred, for example,
subsequent to self-depreciations. ... negative, critical evaluations are
followed by positive, complimentary ones." This suggests that 'self-
depreciation' ('my X is bad') is seen as a typical but not a necessary
aspect of this routine. A more general formula would be:
A: I think X is very bad
B: I don't think the same
I think X is very good
140 Describing conversational routines
2.3. Scaled-down agreements
According to Pomerantz (1978:94), neither upgraded agreements nor
contrastive opposites constitute common responses to compliments.
What does tend to be used in response to a compliment is a 'scaled-down
agreement', and, more specifically, a praise downgrade. For example:
(1) A: I've been offered a full scholarship at Berkeley and at
UCLA.
B: That's fantastic.
A: Isn't that good.
(2) A: Oh it was just beautiful.
B: Well, thank you uh I thought it was quite nice.
The intended generalisation seems to be this: There is (in English) a
common conversational routine which can be represented as follows:
A: I think something [that one can say] about you is very good
B: I think the same
I think it is good
I don't want to say: very good
2.4. Downgrades
Another conversational strategy described by Pomerantz (1978:99) con-
sists in "proposing diminutions of credit", whereby recipients "do not
altogether negate or deny prior assertions but rather downgrade the
prior terms". Two examples:
(1) A: Good shot.
B: Not very solid though.
(2) A: By the way I loved yer Christmas card.
B: I hadda hard time, but I didn't think they were too good...
Here, the generalisation seems to be this:
A: I think your X is very good
B: I don't think the same
I wouldn't say it is very good
because something about it (Y) is not good
'Compliment response' routines 141
Pomerantz (1978:100) comments: "Although these compliment re-
sponses are not contrastive opposites but are rather diminutions and
qualifications of prior praises, they nonetheless are treated as disagree-
ments." The component 'I don't think the same' reflects accurately, I
think, the main point of this comment. As for the 'diminution' and
'qualifications', however, it is hard to be sure if the proposed formula
corresponds to Pomerantz's intentions, because terms of this kind are
so vague and metaphorical that it is by no means clear what exactly
they are meant to stand for.
The same uncertainty applies to the further proposal: "Subsequent to
such disagreements, praise profferers may challenge or disagree with
the diminutions and qualifications and reassert praise." (Pomerantz
1978: 100). This is illustrated as follows:
[1] A: Good shot.
B: Not very solid (though).
A: Ya' get any more solid, you'll be terrific.
[2] A: By the way I loved yer Christmas card.
B: I hadda hard time, but I didn't think they were too good,
but - finally,
A: (Those) were lovely. I thought they were lovely.
Pomerantz (1978: 101) generalises: "Recipients downgrade prior praise,
and profferers upgrade the prior downgrades."
Trying to state what appears to be the intended generalisation, one
could suggest the following:
A: I think your X is very good
B: 1 don't think the same
I wouldn't say it is very good
(because something about it (Y) is not good)
A: I think it (X) is very good
Again, it seems clear that what Pomerantz means by 'downgrades' or
'downgraders' is very different from what, for example, Blum-Kulka -
Danet - Gherson (1985) mean. Yet both she and they regard these
terms as self-explanatory, and do not try to define them. This leads to
confusion, covered up by a semblance of precision. Semantic formulae
of the kind proposed here prevent this kind of confusion.
142 Describing conversational routines
2.5. Reassignment of praise
According to Pomerantz, a common type of compliment responses
consists in 'referent shifts'. These are defined as follows:
AI: A praises B
A
2
: B praises other-than-self
Two kinds of referent shifts are distinguished: 'reassignment of praise'
and 'returns'. The first kind is illustrated with the following example:
A: You're a good rower, Honey.
B: These are very easy to row. Very light.
In exchanges of this kind, "in responding to a compliment, a recipient
may reassign the praise, shifting the credit from himself to an other-
than-self referent" (Pomerantz 1978: 102). Apparently, the intended gen-
eralisation is this:
A: I think that something about you is very good
B: I would want to say something other than this
I would say that something about
something other than me is very good
I have phrased these components in such a way as to avoid any sugges-
tion of disagreement. In particular, I deliberately refrained from saying
'I don't think the same' or even 'I wouldn't say that', because I wanted
to reflect correctly the sense of Pomerantz's (1978:105) comment: "The
compliment response consisting of a second praise (albeit refocused) is
partially supportive of, that is, a partial warrant for or legitimisation
of, the prior praise."
2.6. Returns
'Returns', which are said to be particularly frequent in 'openings and
closings of interactions', are defined as follows:
AI: A compliments B
A
2
: B compliments A
'Compliment responses' in different cultures 143
This is illustrated, among others, with the following examples:
(1) - Ya' sound (justiz) real nice.
- Yeah you soun' real good too.
(2) - Yer looking good.
- Great. So'r you.
These examples suggest the following generalisation:
A: I want to say something good about you
I want you to feel something good because of this
B: I want to do the same to you
I want to say something good about you
I want you to feel something good because of this
It appears that Pomerantz wants to extend the category of 'returns' to
situations when the interlocutors are implying rather than saying good
things about each other; for example:
B: Well anyway nice talking to you
A: Nice talkin to you honey
(Pomerantz 1978: 107)
It seems to me, however, that the formula proposed above does fit
examples of this kind, too: even if the speakers don't say any specific
good things about one another, they both clearly convey that they
would want to say something good about the interlocutor (and would
want to make the interlocutor feel something good because of that); and
that is all that the formula says.
3. 'Compliment responses' in different cultures
Pomerantz opens her study of compliment responses by quoting a letter
from a 'perplexed man' to the Los Angeles Times, together with the
editorial response:
Dear Abby,
My wife has a habit of down-grading sincere compliments. If I say,
'Gee, Ron, you look nice in that dress,' her reply is likely to be, 'Do you
really think so? It's just a rag my sister gave me.' Or if I tell her she did a
144 Describing conversational routines
great job cleaning up the house, her response might be, 'Well, I guess you
haven't seen the kids' room.' I find it hard to understand why she can't
accept a compliment without putting herself down. And it hurts me a little.
How do you explain it, Abby?
Perplexed
Dear Perplexed,
Your wife lacks self-confidence and feels somewhat embarrassed to
accept praise. Don't be hurt. Most people have difficulty accepting com-
pliments with grace.
Abby
A crucial point which is missing from Abby's response, however, is
that responses to compliments differ from culture to culture, and that
within a complex society such as the United States they depend not only
on people's character traits, such as 'lack of confidence', but also on
their cultural background. It is quite possible that 'Perplexed' 's wife did
not lack self-confidence or self-esteem but was simply Jewish, or was
of East European or perhaps Chinese or Japanese background.
More surprising than Abby's lack of attention to this aspect of the
problem, is the fact that Pomerantz herself doesn't mention it once.
Instead, commenting on the letter and the editorial response, she
observes (1978:81): "A large proportion of compliment responses
deviate from the model response of accepting compliments."
But this is a strange observation. If a large proportion of compliment
responses 'deviate from the model' of accepting compliments then is
it not perhaps the model which is inadequate? Shouldn't one rather
speak of a number of different models, operating in different cultures
and subcultures?
The compliment response reported by 'Perplexed' is not a 'deviation
from the model', but a typical instance of a different model, prevailing
in many cultures other than the American WASP culture. Using the
metalanguage proposed here, we could describe the core of the common
conversational strategy which perplexed 'Perplexed', Abby and Pomer-
antz, as follows:
A: I think that something about you (your X) is very good
B: I don't think the same
I think that something about it (my X) is bad
The preferred Anglo-Saxon strategy of what Pomerantz calls 'apprecia-
tion', and which Abby calls 'accepting compliments with grace', can be
represented as follows:
'Compliment responses' in different cultures 145
A: I think that something about you (your X) is very good.
B: Thank you
[i.e.: I know you say this because you want to do something
good for me (you want me to feel something good)
I feel something good towards you because of this]
It should be noted, however, that this 'appreciation' is largely lexical-
ised, as it is in responses to How are you? For example, as in the earlier
case, if one said in response to a compliment Thanks a lot, this would
sound sarcastic rather than 'appreciative'.
Yet another model of compliment responses is characteristic of
Japanese society, and more particularly, of Japanese women's speech.
For example, Mizutani - Mizutani (1987:43) write: "the Japanese ...
will never accept a compliment without saying iie ['no']". This response
pattern is particularly striking in the light of the general Japanese
reluctance to say 'no' under almost any circumstances (cf. Veda's 1974
discussion of sixteen ways to avoid saying 'no' in Japanese). Miller
(1967:289-290) offers the following characteristic example:
Female version:
A: Md, go-rippa na o-niwa de goziimasu wa nee Shibafu ga
hirobiro to shite ite, kekko de goziimasu wa nee
B: lie, nan desu ka, chitto mo teire ga yukitodokimasen mono de
gozaimasu kara, mo, nakanaka itsumo kirei ni shite oku wake
ni wa mairimasen no de goziimasu yo.
A: A, sai de gozaimasho nee Kore dake o-hiroin de goziimasu
kara, hitotori o-teire asobasu no ni datte taihen de
gozaimasho nee Demo rna, sore de mo, itsumo yoku o-teire
ga yukitodoite irasshaimasu wa. ltsumo honto ni o-kirei de
kekko de goziimasu wa.
B: lie, chitto mo sonna koto goziimasen wa.
A: 'My, what a splendid garden you have there - the lawn is
so nice and big, it's certainly wonderful, isn't it!'
B: 'Oh no, not at all, we don't take care of it at all any more, so
it simply doesn't always look as nice as we would like it to.'
A: 'Oh, I don't think so at all - but since it's a big garden, of
course it must be quite a tremendous task to take care of it
all by yourself; but even so, you certainly do manage to make
it look nice all the time; it certainly is nice and pretty any
time one sees it.'
B: 'No, I'm afraid not, not at all.'
146 Describing conversational routines
According to Miller (1967:290), a male version of the same dialogue
could look as follows:
Male version:
A: Ii niwa da nii?
B: Un.
A: 'It's a nice garden, isn't it?'
B: ['Mm.']
The patterns underlying these two exchanges can be represented
as follows:
Female version:
A: I think something about you (your X) is very good
B: I don't think this
I think it is not good
A: I don't think this
I think it is good
B: I don't think this
I think it is not good
Male version:
A: I think this X is good
B: I think the same
What applies to compliments (especially in women's speech) applies
also to praise, especially to praise directed at personal achievements.
Thus, Mizutani and Mizutani write:
Except among good friends, the Japanese usually deny any praise received
from others.... To deny praise of one's skills or abilities, one will say:
lie, madamada-desu. (No, I'm not any good at it yet. - lit. No, not yet.)
lie, watashi-nanka dame-desu. (Oh, I'm so poor at it. - lit. Such a person
as me is no good.)
(Mizutani - Mizutani 1987:43)
This pattern can be represented as follows:
A: I think you did something very good
B: I wouldn't say this
I am not good
I can't do things well
Conclusion 147
If one's achievements are absolutely undeniable and praise cannot
be simply rejected, one will use a slightly different pattern:
Maa, nantoka.
'I manage to do it somehow.'
Okagesama-de nantoka.
'Thanks to everybody I could manage.'
Doo-yara koo-yara.
'Somehow or other I could.'
(Mizutani - Mizutani 1987:43)
Similarly, according to another account:
In daily life when being praised by others, the person praised often denies
the merit by saying something like, 'I don't deserve the praise'. Otherwise,
after expressing his/her thanks for the praise, the person adds some words
in which he/she emphasises good luck, a favour from others, or the help of
his/her surroundings. By so doing, he/she tries to ascribe his/her virtue or
achievements to the power of something other than his/her own ability
or efforts. (Honna - Hoffer 1989:240)
This pattern can be represented as follows:
A: I think: you did something very good
B: I know:
it is not because I am good
(I am not good)
I believe that when different conversational routines are represented
in this way, that is, if they are modelled in a controlled semantic
metalanguage (derived from natural language and yet largely language-
independent), they can be clearly identified and easily compared, both
within and across language and culture boundaries.
4. Conclusion
Empirical analyses of conversational interaction of the kind undertaken
by Pomerantz are, potentially, very important from the point of view of
cross-cultural studies, both theoretical and applied. For example, it is of
course very important for the immigrant to know what the rules of the
148 Describing conversational routines
different conversational games in the new country are, and how they
differ from those prevailing in the old country. But comparisons of this
kind cannot be carried out without a culture-independent descriptive
framework, and in particular, without a culture-independent semantic
metalanguage.
In his Introduction to Studies in the organisation of conversational
interaction, the volume containing Pomerantz's study, Schenkein
(1978:3) wrote: "the descriptions presented here offer promising move-
ment towards an empirically based grammar of natural conversation". I
am all in favour of an empirically based study of natural conversation,
and I regard Schenkein's volume in general, and Pomerantz's study
in particular, as valuable, interesting and important. But to be truly
fruitful, empirical studies of course require a well-justified theoretical
framework, and adequate descriptions require not only reliable data but
also a justifiable metalanguage.
I suggest that categories such as 'upgrade', 'downgrade', 'acceptance'
or 'rejection', 'supportive action', 'contrastive opposites', or 'returns'
may be useful at some stage of analysis, but cannot be relied on as
adequate analytical tools. They do not provide an adequate framework
for describing and comparing conversational routines within anyone
culture (for example, in the mainstream middle-class white American
culture); and they are even less adequate as a framework for comparing
the organisation of conversational interaction across cultural boundaries.
Here as elsewhere in cross-cultural analysis, what is needed is
language-independent and 'culture-free' analytical tools; and these can
be found neither in English folk-categories such as compliments and
credits, agree and disagree, or accept and reject, nor in arbitrarily
invented vague and metaphorical labels such as upgrade or downgrade;
but in universal (or near-universal) human concepts such as 'good' and
'bad', 'want' and 'know', 'think' or 'say' - that is to say, in the same
'natural semantic metalanguage' in which meanings encoded in different
linguistic systems can be described and compared.
Chapter 5
Speech acts and speech genres across languages
and cultures
1. A framework for analysing a culture's 'forms of talk'
Every culture has its own repertoire of characteristic speech acts and
speech genres. As Baxtin (1979:257) pointed out, in speaking as in
writing, "we 'pour' our speech into ready-made forms of speech genres
... These forms are given to us in the same way in which our native
language is given."3
Following Baxtin, I regard it as essential that complex speech genres
and 'speech events' (such as lecture, letter, or gossip) should be treated,
on some level, in the same way as simple speech acts (such as question,
request, promise, or warning). This is not to deny that terminological
and conceptual distinctions such as that between 'speech genre', 'speech
event', and 'speech act' may be useful in certain contexts. But it is also
important to stress that despite the tremendous variety of speech genres
(using this term in Baxtin's sense, as a cover-all term) - in function,
structure, and above all in length - they share, in an important sense,
the same linguistic nature and require a unified descriptive framework.
(The need for making such conceptual distinctions and yet studying the
whole range of phenomena in question within a unified framework
was pointed out in Hymes 1962.)
The idea that different cultures can be studied and compared via their
characteristic speech genres has by now become widely accepted,
although it is seldom recognised that this statement applies to simple
speech acts as much as to complex speech genres. It seems to me, how-
ever, that the cross-cultural study of speech genres has been severely
hampered by the absence of a culture-independent and language-
independent semantic metalanguage.
150 Speech acts and speech genres across languages and cultures
1.1. The importance of folk labels
I believe that one particularly fruitful way of approaching the repertoire
of speech acts and speech genres characteristic of a given culture is via
their folk names, that is, the special lexical units which have come to
encode a culture's view of its most relevant 'forms of talk' (cf. Goffman
1981). This is not to say that all 'language games' (cf. Wittgenstein
1953) played in a given culture will necessarily have their own folk
names. Nonetheless, it seems reasonable to assume that, generally
speaking, those language games which do have such names are more
relevant to a given culture than those which don't.
In Gumperz' s (1972: 17) words, "members of all societies recognise
certain communicative routines which they view as distinct wholes,
separate from other types of discourse, characterised by special rules
of speech and nonverbal behaviour." Crucially, "these units often carry
special names" (1972: 17). Consequently, "one good ethnographic tech-
nique for getting at speech events, as at other categories, is through
words which name them" (Hymes 1962: 110).
1.2. Two approaches
Speech genres characteristic of a given culture are usually described in
one of two ways: either from outside or from inside. If they are described
from outside, researchers raise problems such as 'questions in Eskimo',
'commands in Zulu', or 'blessings and curses in Yakut'. If they are
described from within, we read about speech genres such as namakke,
sunmakke, kormakke in Cuna or rapping and capping in Black English
(cf. Sherzer 1974; Abrahams 1970).
The dangers of the first approach seem to be evident. English words
such as question, command or blessing identify concepts which are
language-specific. They embody an English folk taxonomy, which,
like all folk taxonomies, is culture-specific. To approach a repertoire of
non-English speech genres through a grid of English folk concepts
means to risk a biased and ethnocentric description.
On the other hand, if one describes speech genres characteristic of a
culture from 'within', in terms of concepts such as namakke or rapping,
the results may be free of any ethnocentric bias, but they risk being
somewhat hard to grasp to the outsiders. Of course, studies such as
Sherzer (1974) or Abrahams (1970) do a great deal towards explaining
A framework for analysing a culture's 'forms of talk' 151
the meaning and the illocutionary force of the relevant genres to the
outside world, but the fact that these explanations take the form of
lengthy studies and that they are never summed up in succinct and
intuitively comprehensible formulae, constitutes, it seems to me, a
serious obstacle to genuine understanding.
The same applies of course to the first approach. Authors of studies
such as 'questions in Eskimo' or 'commands in Zulu' or 'insults in Black
English' often take great pains to explain that the speech acts or speech
genres in question are not really 'questions', 'commands', or 'insults' in
our sense of the term, and that they use these English terms only for
convenience's sake. It seems to me, however, that since the explanations
of what these genres really are also take the form of long studies
which are never summed up in succinct and intuitively comprehensible
formulae (other than the nearest English folk labels), it is very hard for
the reader to grasp their culture-specific essence.
To my mind, two things have to be recognised at the same time: First,
folk names of speech acts and speech genres are culture-specific and
provide an important source of insight into 'communicative routines'
most characteristic of a given society; and second, to fully exploit this
source, we must carry out rigorous semantic analysis of such names and
express the results of this analysis in a culture-independent semantic
metalanguage.
Searle has claimed that:
Illocutionary acts are, so to speak, natural conceptual kinds, and we should
no more suppose that our ordinary language verbs carve the conceptual
field of illocutions at its semantic joints than we would suppose that our
ordinary language expressions for naming and describing plants and ani-
mals correspond exactly to the natural biological kinds. (Searle 1979:ix-x)
I would insist, however, that when philosophers of language discuss
illocutionary acts such as 'promising' (cf. Searle 1969:54-71), 'request-
ing', or 'commanding' (cf. Searle 1979:3), they are not discussing any
language-independent 'natural kinds'. They are discussing complexes
of components which have been singled out from among countless
combinations of components occurring in human communication by
concepts which are English-specific.
It seems fairly obvious that more or less ritualistic speech acts, such
as 'baptising', 'exorcising', or 'absolving sins', cannot be viewed as
culture-independent natural kinds, analogous to biological natural kinds.
In my view, it is an illusion to think that acts such as 'promising', 'order-
152 Speech acts and speech genres across languages and cultures
ing', or 'warning' are any less culture-dependent (cf. Rosaldo 1982; see
also Verschueren 1985).
The authors of a recent study (Fraser - Rintell - Walters 1980:78)
say that their research is based on the following assumption: "Every
language makes available to the user the same basic set of speech acts,
such as requesting, apologising, declaring, and promising, with the
exception of certain culture-specific ritualised acts such as baptising,
doubling at bridge, and excommunicating." Underlying this assumption
is "the claim that if one language permits an act such as requesting,
every other language will. Though there may be certain exceptions as
one moves from the basic everyday acts such as requesting to the more
culture-specific ones such as baptising, exceptions to this claim have not
arisen and do not appear likely to, given what we know today about
language" (1980:79).
In my view, it would be difficult to base one's research on a less
justified assumption.
1.3. Some examples: English vs. Japanese
To focus on one example, a concept such as warning is not as language-
independent as the concept of mimosa pudica or felis domesticus is;
rather, it is as language-specific as the concept of shrub or bug is. It is
a function of the meaning of the English word warning, not a God-given
or science-given Urdatum. English speech-act verbs codify a folk
taxonomy of speech acts, not some culture-independent, scientific or
philosophical taxonomy of modes of human communication.
There are many languages which have no exact equivalent of the
word warning and which have, instead, words for modes of communica-
tion which have no equivalents in English. For example, Japanese has
the word satosu, which combines some of the components of the
English concept codified in the word warning with some other compo-
nents: an assumption that the speaker has authority over the addressee,
the intention of protecting the addressee from evil, and good feelings
toward the addressee (see Nevile 1981). In English, the assumption of
authority is encoded in verbs such as order and forbid, but it is never
combined (lexically) with the intention to protect. This combination of
components: 'I am your superior; I am responsible for you; I don't want
you to do anything bad; I care for you', is a characteristic feature of
Japanese culture, in which the relationship between a superior and a
A framework for analysing a culture's 'forms of talk' 153
subordinate is likened to that between a parent and a child (see Nakane
1970, 1972; Lebra 1976; Smith 1983), and this feature is reflected in the
meaning of the Japanese verb satosu. The fact that English doesn't
have any verb which would combine authority, responsibility, and care
seems also significant.
Using the semantic metalanguage based on universal semantic primi-
tives we can show both the similarities and the differences between
different speech acts clearly and explicitly. For example, the relationship
between warn and threaten can be represented along the following lines
(cf. Wierzbicka 1987):
warn
I say: if you do X something bad (Y) may happen to you
I think: if you know it you may not do X
I say this because I want you to know it
threaten
I say: if you do X I will do something bad (Y) to you
I think: if you know it you may not do X
I say this because I want you not to do X
The Japanese concept encoded in the word satosu can be represented
along the following lines (cf. Nevile 1981; Morimoto 1985):
satosu
(a) I say:
you should do X
(b) it will be bad if you don't do it
(c) I say this because I think I should say it
(d) I think:
you will do it because of this
(e) you know that I feel something good towards you
(f) I know:
I can say things like this to you
(g) you can't say things like this to me
Component (a) of this explication shows that satosu is similar, in some
respects, to the English concept of advise; (b) links it with the English
concept of warn and admonish; (c) shows that the speaker feels respon-
sible for the addressee's actions; (d) shows the speaker's confidence in
his own influence; (e) spells out the source of this confidence, and of the
154 Speech acts and speech genres across languages and cultures
speaker's interest in the addressee's actions; while (f) and (g) spell out
the asymmetrical character of the relationship.
The asymmetrical character of satosu may seem to link it with such
power-based English words as order or reprimand. But satosu does not
refer to the speaker's power over the addressee. Rather, it seeks its
legitimisation in the speaker's resonsibility for, and good feelings
towards, the addressee, that is, in a parent-like attitude of the superior
(speaker) towards the subordinate (addressee). Thus, the concept of
satosu is a clear manifestation of the much-discussed Japanese 'paternal-
ism'. In Japan:
The boss ... is supposed to have parent-like feelings. Indeed the social
expectations of his role often cause him to display overt behavior
suggesting such feelings whether they are present in him or not. There
are indeed positive fantasies of being protected directed towards superiors.
. .. This interplay of expectations resembles the expectations of increased
responsibility directed towards the eldest son in the traditional family. He
is to internalise the sense of responsibility for others under his authority.
Responsibility for taking care of personal matters would be considered
intrusive by subordinates were they to be exercised in the west. For
example, business executives, foremen, or even higher executives in
Japan will sometimes act as go-betweens in assuring a proper marriage
for one of their subordinates. This is seen as a part of a parent-like respon-
sibility and indeed a type of nurturant concern with the subordinate.
(DeVos 1985:159)
As Lebra (1976:51) points out, in prewar Japan even "military units
formed pseudofamilies consisting of pseudoparents and pseudochildren.
A former officer is quoted as saying, 'The warrant-officer is like a house-
wife who takes good care of soldiers as a mother, while the company-
commander may be likened to a father whose orders are strictly observed
but who has the affection of kinship towards his soldiers'." (Lebra is
quoting here from Minami 1953:157).
If the superior is expected to behave like a parent, the subordinate is
expected to assume the role of a child - not so much an obedient
child, as a child who knows and can rely on the parent's affection. This,
too, is reflected in the lexicon of Japanese speech-act verbs, and most
particularly (as pointed out by Morimoto 1985), in the concept of
nedaru. But first a couple of additional quotes from the work of social
psychologists:
A framework for analysing a culture's 'forms of talk' 155
The 'child' -role player can expect to depend upon the 'parent' -role player
for security and protection by appealing to the latter's oyagokoro ('parental
sentiment'), which is characterised as warm, benevolent, and nurturant.'
(Lebra 1976:51).
In the traditional Japanese system there were no 'rights' on the part of the
subordinate. The only recourse for subordinates in the past, since they
had no contractual relationships, was to hope to induce kindness and
benevolence in their superiors. These feelings were induced by invoking
potential feelings of nurturance and appreciation from them. This capacity
to induce kindness and benevolence in superiors in a manipulative manner
is called amaeru in Japanese. It has been very cogently discussed at length
by Takeo Doi [1973]. (De Vos 1985:159-160)
As pointed out by Morimoto (1985), Japanese has two verbs usually
translated into English as 'ask (for)': tanomu and nedaru; but while
tanomu can indeed be regarded as a semantic equivalent of ask, nedaru
implies a rather different, characteristically Japanese concept, which
reflects the psychology of amae. (For a semantic analysis of the concept
of amae/amaeru, see Wierzbicka, to appear, chap. 4.)
Nedaru invariably implies intimate relationship and I think there is some-
thing more than 'I assume you will want to do it' about this verb. It is
used only by a subordinate to a superior with whom they have a close
relationship ... and it shows that a speaker knows that a superior has a
friendly feeling towards him and that he could take advantage of it.
(Morimoto 1985)
Drawing on Morimoto's analysis of nedaru, I would propose the follow-
ing explication of this concept:
nedaru
(a) I say: I want you to do something good for me
(b) I say this because I want you to do it
(c) I know: you don't have to do it
(d) I think you will want to do it
(e) because I know that you feel something good towards me
(f) I know: you can do good things like this for me
I can't do good things like this for you
Both the English ask (for) and the Japanese tanomu include the first
three of these components (a, b, and c), but they do not include (d), a
confident expectation that one's wish will be granted because of the
addressee's presumed benevolence; (e), an assumption of the addressee's
156 Speech acts and speech genres across languages and cultures
'good feelings'; or (f), an assumption that the relationship is inherently
asymmetrical, not so much in terms of power and authority as in terms
of nurturance and dependence.
In the formulae proposed here, the meaning of both the English verbs
warn and threaten and of the Japanese word satosu have been repre-
sented in a kind of simplified and standardised English. Nonetheless,
these formulae could be readily translated into Japanese, because while
words such as warn (or abuse, swear, or curse) do not have Japanese
equivalents, semantically simpler words, such as I, want, know, do,
happen, bad, or because, do have equivalents - not only in Japanese but
also in most other languages of the world. For this reason, the semantic
metalanguage used here can be regarded as, in a sense, language-
independent.
Coulmas (1981 :70) writes: "The difficulty boils down to the general
question of how speech acts can be cross-culturally compared and 'trans-
lated'. To treat speech acts such as thanks and apologies as invariable
abstract categories is surely a premature stance." I would put it more
strongly: It is not merely premature, it is downright ethnocentric.
Coulmas (1981:81) elaborates: "After all, 'thanks' and 'apology' are
Western words ... But the applicability [to other cultures] of such cate-
gories should not be taken for granted. In particular, we should not
assume that names of speech acts of individual languages define univer-
sal types of speech acts." I couldn't agree more. Unfortunately, Coulmas
continues, "With this in mind we can now approach the problem of the
cross-cultural comparability of thanks and apologies. ... As regards
apologies and thanks, it seems to be a reasonable assumption that they
exist as generic speech acts in every speech community" (1981:81).
But in fact, Coulmas himself convincingly shows that the concepts
encoded in the English words thanks and apology don't really fit
Japanese culture. Shouldn't one firmly conclude, therefore, that they are
not suitable tools for cross-cultural comparison of speech acts? They
may serve a purpose as long as one is trying to show that they don't fit
cultures such as Japanese. We can also ask legitimately: What do the
Japanese do in those situations in which we would apologise or thank?
But to describe speech acts which are characteristic of Japanese culture
in positive terms, one needs a metalanguage which would not be derived
from English speech-act labels such as thank or apologise.
To thank someone means, roughly, to say that we feel something good
towards them because of something good they have done for us. To a
European, this may seem a perfectly 'natural' and 'normal' response to a
A framework for analysing a culture's 'forms of talk' 157
favour received. But in Japanese culture, with its stress on social
hierarchy, and above all, with its stress on obligatory repayment of all
favours, this response is much less 'normal' and generally applicable. As
Coulmas himself points out, in Japanese culture it is 'normal' to respond
to other people's favours in the same way as one responds to one's
own transgressions against other people: that is, by saying that one
feels 'something bad' rather than 'something good' because of this. Thus,
expressions such as sumimasen, literally, 'it never ends' (i.e., I am aware
of my 'never-ending indebtedness' to you) are used both in situations
calling, from a European point of view, for thanks, and in situations
calling for an apology.
It is not surprising, therefore, that Japanese doesn't have a verb
corresponding to the English verb to thank. The closest word it has is
kansha suru (of Chinese origin), but this could never be used with
respect to a child thanking her mother for a present, or to a lecturer
thanking her student for some service, and in fact, it is virtually
restricted to written language (Kaoru Sakurai, p.c.). The idea that
regardless of the status, rank, and the type of relationship one can
always react to a favour in essentially the same way is alien to Japanese
culture, and the absence of a general speech-act verb corresponding to
thank reflects this. By translating both the English word thank and the
Japanese word kansha suru into the metalanguage of universal semantic
primitives we can reveal both the similarities and the differences
between them, and in this way we can document cultural differences
which may otherwise seem elusive and non-accessible to rigorous
analysis. I propose the following (for a slightly different explication
of thank, and for discussion, see Wierzbicka 1987):
thank
(a) I know: you did something good for me
(b) I feel something good towards you because of this
(c) I say this because I want you to feel something good
kansha suru
(a) I know: you did something good for me
(b) I say: I feel something good towards you because of this
(b') I know: I couldn't do something good like this for you
(b") I feel something bad because of this
(c) I say this because I think I should say it
158 Speech acts and speech genres across languages and cultures
Components (a) and (b) of these explications are the same; but the
'illocutionary purpose' (c) is different in each case, with the Japanese
concept stressing a feeling of obligation on the part of the speaker. Most
importantly, however, the Japanese concept implies an asymmetrical
relationship (component b'), and a feeling of 'unrepayable debt', which
links the feeling of quasi-gratitude with something like a feeling of
guilt (component b").
It should be added that the expectation that one would find 'generic
speech acts' such as thanks and apologies in most (if not all) cultures is
completely unfounded. For example, among the Australian Aboriginal
Yolngu people:
Yolngu never express 'thanks' verbally to each other ... people normally
do things for one of two reasons: either because they want to, or else
because they have some obligation to fulfil to specific relations. Therefore,
the yolngu for whom a balanda [white person] has just done a 'favour',
automatically thinks that the balanda gave him the lift in his boat because
he wanted to, so there is no need for any expression of thanks.... This
analysis of the two reasons why yolngu do anything, is quite consistent
with the principles of reciprocity ... because the principle of reciprocity
acts within the system of obligations to specific relations. (Harris
1984: 134-135)
On the other hand, kinship-based obligations, which in part explain
the absence of 'thanks' in many Australian Aboriginal languages, are
reflected in other speech acts, which have no counterparts in European
languages, and which have no place in the lexicalised taxonomies of
speech acts in a language like English. This point will be considered
more closely in the next section.
1.4. Another example: English vs. Walmatjari
Every speech act or speech genre constitutes a bundle of illocutionary
components: expressed intentions, assumptions, thoughts, feelings, and
so on (see also Chapter 6). For example, the act identified in the English
folk taxonomy as order (as in 'X ordered Y to do Z') contains, I suggest,
the following semantic structure:
order (I order you to do X)
(a) I say: I want you to do X
(b) I say this because I want you to do it
A framework for analysing a culture's 'forms of talk' 159
(c) I think you have to do it because of this
(d) I think you will do it because of this
The act identified in English by means of the verb ask (as in 'X asked Y
to do Z') can be assigned the following semantic structure:
ask
(a) I say: I want you to do something good for me (X)
(b) I say this because I want you to do it
(c) I think: you don't have to do it
(d) I don't know if you will do it
But in the Australian language Walmatjari, spoken in Western
Australia, there is another characteristic speech act, related to both the
English ask (tolfor) and the English order, but equivalent to neither (see
Hudson 1985). It is based on kinship rights and obligations, which are
so characteristic of Aboriginal society. Thus, Walmatjari has a special
word, japirlyung, for what might be called 'kinship-based requests', the
idea being that a kinship-based 'request' cannot be refused. But in call-
ing this speech act a 'request', one is, of course, committing an error:
an act which, in the speaker's view, cannot be refused is not really a
'request'. To represent this act in a way which would be free of an
Anglo-centric bias and yet which would be meaningful to an outsider,
one can, I suggest, use universal (or near-universal) elementary (or
near-elementary) semantic units, along the following lines:
japirlyung
(a) 1 say: 1 want you to do something good for me (X)
(b) I say it because I want you to do it
(c) 1 think: you have to do good things for me
(c') I think you know: everyone has to do good things for some
other people (because of the way we are related)
(d) I think: you will do it because of this
Components (a) and (b) of this explication are the same as components
(a) and (b) of the English verb ask (for). But component (d) expresses a
confidence that the addressee will comply, associated in English not with
ask but with order. The English ask, on the contrary, implies that the
speaker doesn't know whether the addressee will comply: 'I don't know
whether you will do it' - a component which links ask (tolfor) with
questions. But the basis of confidence is quite different in the case of
japirlyung than it is in the case of order. Order implies a hierarchical
160 Speech acts and speech genres across languages and cultures
relation: if I say to you that I want you to do something then you have to
do it because of this. Japirlyung implies nothing of the sort: it doesn't
imply that the addressee is under the speaker's authority or control; it
only implies that the addressee has the obligation to do some 'good
things' for the speaker because of the way they are related, and it recalls
a general system of kinship-based obligations, a system which involves
everyone in the community. (For a semantic analysis of the concept
of 'relatedness', see Wierzbicka, to appear, chap. 9.)
Hudson (1985:71) offers the following example of the use of
japirlyung.
Japirlyinya parla parri-ngu nganpayi kuyi-purru.
asked he-him boy-ERG man meat-PURP
'The boy asked the man to give him meat (because of kinship
obligations). '
She contrasts this with the use of the verb jinjinyung, which, like the
English order, "implies the authority on the part of the speaker rather
than obligation based on kinship" and which "does not include the
component 'that is good for me', which is present in japirlyung". For
example:
Jinjinyinya manya yinparnu-purru.
ordered he-them sing-PURP
'He ordered them to begin singing the corroboree.'
As Hudson points out, in this last example "the speaker has authority
in the corroboree, and so he is able to order the performers to begin"
(1985:71).
Thus, the idea of a 'directive' based on authority is present (lexical-
ised) in both English and Walmatjari; on the other hand, the idea of
a 'directive' based on kinship-obligations is present (lexicalised) in
Watmaljari but not in English. It might be added that the idea of a
directive based neither on authority nor on kinship-obligations (but on an
appeal to individual good will) is lexicalised in English but apparently
not in Walmatjari. This fits in well with Harris' remarks on the absence
of the concept of 'thanks' in Aboriginal culture.
A framework for analysing a culture's 'forms of talk' 161
1.5. The elimination of vicious circles
The use of relatively simple terms in the explications, which makes
possible a rigorous comparison of concepts (both within one language
and crosslinguistically) ensures also the elimination of the vicious
circles which have plagued traditional dictionaries in general and
dictionaries of synonyms and related words in particular. Traditionally
(as pointed out in Chapter 1 above), a word like request has nearly
always been described with reference to ask for, and vice versa (along
the lines of 'to request' = 'to ask politely for something' and 'to ask for
something' = 'to make a simple request'). For example, Hornby et al.
(1969) define ask, in the relevant sense, as "request information or
service", request as "make a request", and request (n.) as "asking or
being asked for something".
In my analysis, no speech-act verb can be defined in terms of another
speech-act verb. The only verb referring to speech which can occur in
the explications is say, which I regard as indefinable and which has the
status of a universal semantic primitive. The other words used in my
explications do not always have this status, but they are all relatively
simple, and for present purposes can be treated as indefinable. The strict
separation of the words which are being defined from the small set of
relatively simple words which are used for defining prevents vicious
circles and ensures that the analysis is real, in Aristotle's sense, that
is, that it consists in reducing posteriora to priora (complex and
relatively obscure concepts to simpler and relatively clear concepts), not
in translating unknowns into unknowns.
1.6. Evidence for the proposed formulae
The formulae proposed should predict correctly the entire range of a
term's use. Usually, each formula goes through a great many versions,
being checked and rechecked with native speakers, before an optimal
version is reached - optimal from the point of view of accounting for
all the aspects of the term's use.
Accounting for the entire range of use means, for me, accounting for
both the situations to which the term would be applied and the syntactic
environments in which it would be used.
Consider, for example, the verbs reveal, confess, and tell. Why can
one tell someone a joke but not *reveal someone a secret or *confess
162 Speech acts and speech genres across languages and cultures
someone one's sins? Presumably, in all three cases, the addressee is
affected by the action (in that the addressee comes to know something).
But reveal and confess differ from tell in their implications with regard
to the message. Revealing a secret to somebody crucially affects the
secret, as well as the addressee, because the secret comes into the open.
Similarly, if I confess my sins to somebody, my sins cease to be secret,
they cease to be a burden on my soul, they may even be 'washed away'
by an act of absolution. It is quite different with tell. Tell doesn't imply
that the message is secret, as reveal does, and it doesn't imply that the
message is a guilty one, as confess does.
The message of tell can be quite trivial, and it need not to be affected
by the speech act. For example, if I tell John a joke, it is likely that this
will affect John (he will laugh), but it will not affect the joke. And even
if the message is not trivial, for example, if I tell John 'the truth', John
will probably be affected, because he will come to know the truth, but
there is no implication that 'the truth' in question will be affected, for
example, in the sense of becoming public or coming into the open, as
a secret does when it is revealed.
The relative unimportance of the addressee in comparison with the
message implied by verbs such as reveal or confess is reflected in the
ability of the addressee phrase to be omitted, as well as in its inability
to be 'promoted' to the internal dative position:
He revealed that he had spent ten years in jail.
He confessed that he had spent ten years in jail.
??He told that he had spent ten years in jail.
In devising semantic explications for the verbs reveal, confess, and
tell I have tried to differentiate the formulae accordingly. In this way,
semantic analysis is used to explain differences and similarities in
syntactic patterning, and syntactic patterning is used as evidence for
semantic explications. (For further data and discussion, see Wierzbicka
1988, chap.6.)
1.7. The first-person format
One last feature of the analysis offered in the present chapter requires
a comment: the first-person (singular, present tense) format of the
explications. This first-person mode of analysis is a major and deliberate
departure from generally accepted conventions. It reflects my conviction
A framework for analysing a culture's 'forms of talk' 163
that the first-person form of speech-act verbs ('I warn', 'I request', etc.)
is semantically simpler than all the other forms, and that expressions
such as 'he warned', 'he requested', and so on, are semantically derived
from the corresponding first-person expressions (or from their para-
phrases). This is so despite the fact that some speech-act verbs (such as,
for example, boast or threaten) are never used in the first-person present
tense (in a performative sense). I believe that reports such as 'John
boasted of X' or 'John threatened to do X' rightly or wrongly attribute
to the author of a given speech act a subjective attitude which can
only be described adequately in the first-person mode.
To see this, it should be enough to consider the fact that the subjective
attitudes in question are often expressed simply by intonation (cf. Deakin
1981) and that intonation has inherently a first-person meaning. The
intonation can convey 'I am angry', but never 'she is angry'; it can
convey 'I want you to do something' or 'I want you to tell me some-
thing', but never 'she wants him to do something' or 'she wants him to
tell her something'.
To describe the meaning of illocutionary forces in a third-person for-
mat is tantamount to deriving direct discourse from indirect discourse.
Yet it is well known that many languages have no indirect speech and
that in children's speech, utterances such as 'Daddy said: what time is
it?' 'Mummy said to him: I love you' occur earlier than utterances
such as 'Daddy asked what time it was'. 'Mummy told him that she
loved him' (cf. Coulmas 1986).
1 am not suggesting that all verbs which can be used in reporting
speech have an inherent first-person perspective. For example, verbs
such as ejaculate, bark, babble, fume, thunder, mutter, or snap describe
speech events from outside, spelling out the impression of an external
observer. One would be unlikely to use them in the first person, even
in the past tense:
'Too late', she barked / ?/ barked.
'/ absolutely forbid it', thundered the Colonel / ?/ thundered.
Verbs of this kind describe the manner of speech, and while they do
so partly in terms of emotions which one would assume they would be
associated with, they imply nothing about the speaker's illocutionary
purpose. They can be roughly paraphrased in the third person, for
example:
164 Speech acts and speech genres across languages and cultures
"X", he thundered. = "X", he said, saying it in a way which
would remind one of the sound of thunder, as people do when
they want to show that they are angry and that they have power.
What real speech-act verbs (complain, boast, warn, order, promise,
announce, etc.) imply about the manner of speech is that it must be com-
patible with the attitude attributed to the speaker; and they directly
attribute to the speaker an attitude which can be accurately portrayed
only in terms of a first-person perspective. (For a different view on this
point, see Boguslawski 1988.)
1.8. The problem of other minds
The first-person format of the analysis solves the paradox that in the
view of some linguists (for example, Chomsky 1975) makes all semantic
analysis of speech acts a futile enterprise. It is quite obvious that
speech acts differ from one another in terms of the speakers' subjective
attitudes, that is, in terms of their assumptions, intentions, and so on.
However, other people's assumptions, intentions, and so on, cannot be
observed and ultimately remain unknown to outsiders. How can we then
develop a rigorous semantic analysis on the basis of these unknowns?
I regard this objection as valid, but only with respect to the conven-
tional third-person format of speech-act analysis. If John warns some-
one about something, nobody other than John knows what John's real
intentions are. This, however, does not detract from the validity of
the equation:
I warn you =
I say: if you do X something bad (Y) may happen to you
I say this because I want you to know it
I think if you know it you may not do it
In explicating speech-act verbs in a first-person format we are modelling
the attitudes conveyed in first-person expressions (such as 'I warn
you') or attributed to the speakers, rightly or wrongly, in third-person
reports (such as 'he warned'). Whether the assumptions and intentions
expressed in our formulae are sincerely held by people who perform
appropriate speech acts is, from a semantic point of view, quite irrele-
vant. I am not claiming anything about the real intentions of a person
who warns, threatens, or requests. I am only claiming that when some-
Some Australian speech-act verbs 165
one says I warn you or Careful! This gun is loaded!, the attitude
conveyed can be described in the verb warn, and that when someone
says he warned me, that attitude is attributed to the speaker (cf. Skinner
1970; Hymes 1974b:182-183). Similarly, one can very well say:
She ordered him to go, but she didn't really want to be obeyed.
This means that it would be incorrect to describe a sentence such as 'X
ordered Y to do Z' in terms of 'X wanted Y to do Z'. But it would be
correct to describe it in terms of X's saying (or otherwise conveying
the meaning of): 'I want you to do Z' (cf. Wierzbicka 1974).
Semantics is not concerned with people's 'real' (as opposed to con-
veyed) intentions and assumptions. The task of speech-act analysis
consists in modelling in explicit and verifiable formulae the attitudes
that people convey in speech by conventional linguistic means (which,
of course, include the intonation). The task of semantics in general con-
sists in modelling in explicit and verifiable formulae the meanings which
people convey in speech (again, by conventional linguistic means).
In what follows, I am going to discuss a number of language-specific
speech acts and speech genres, drawn from a few different languages,
trying to show that each of them embodies a mode of social interaction
characteristic of a particular culture. Informal discussion will be supple-
mented by rigorous semantic description, which will take the form of
explications formulated in the proposed metalanguage of universal
semantic primitives. I will start with a fairly extensive discussion of
five speech-act verbs which have emerged as key concepts in Australian
English. This will be followed by a more summary analysis of a few
concepts which stand for more complex speech genres: the black English
'dozen', the Hebrew 'dugri talk', and the Polish kawal and podanie.
2. Some Australian speech-act verbs
2.1. Chiack (chyack)
The Australian English word chiack ['(t)fajrek], allegedly derived from
"the cockney pronounciation of 'cheek' - impudent badinage" (Bulletin
1898, in AND), refers to a characteristically Australian form of social
interaction and reflects a characteristically Australian form of humour.
166 Speech acts and speech genres across languages and cultures
(The word is highly colloquial and since it belongs, essentially, to spo-
ken rather than written language, its spelling is variable.) Essentially,
'chiacking' consists in saying something bad about the addressee for
shared fun. Australians themselves are inclined to see 'chiacking' as
one of their favourite national pastimes, and forms of entertainment.
Most examples of this word cited by either Wilkes (1978) or AND (1988)
refer in fact to a habitual, rather than occasional, activity of 'chiacking'.
In examples from these sources, Dawes (1943, in AND) talks about the
"Australian passion for handing out chiack", and Hardy's reference
to the old chiack indicates that 'chiacking' is very much part of the
familiar (and hence positively viewed) Australian way of life:
Hullo, hullo, Chilla said, always a bit too keen on the old chiack,
especially when it came to Tich's unsuccessful carryings on with
the female of the species. (Hardy 1971) (AND)
Other characteristic examples include the following:
My mates chyacked me all night. (Australasian Printer's Keep-
sake, 1885) (AND)
Diggers of the Yarra tribe ... like to chiack the Cornstalk variety
about 'our' arbour'. (Aussie 1919) (AND)
They whooped, they made ribald noises, they chyacked one
another. (S. Campion 1944) (AND)
They chyacked their sissy mates and their sisters who were
forced to attend late afternoon dancing classes. (R. McKie 1977)
(AND)
As these examples indicate, 'chiacking' is closely associated with the
Australian idea of 'mateship': it is usually done among 'mates', and it
is often done reciprocally, and if not reciprocally among mates, then
collectively with mates. Usually, the men speak one at a time, making
negative remarks about the addressee, while the other men are laughing,
so that a group of 'mates' constitutes both a group of participants and
an audience, as in the following examples:
The circle offrivolous youths who were yelping at and chy-acking
him. (Australian Monthly Magazine 1879) (AND)
They're always a-poking borack an a-chiackin' 0' me over in the
hut! (l.A. Barry 1893) (Wilkes 1978)
Some Australian speech-act verbs 167
There were several pretty girls in the office, laughing and chiack-
ing the counter clerks. (Henry Lawson 1896) (Wilkes 1978)
Don't walk about; it's tirin'; stand at street-corners and spit -
besides that ther best place ter see life and chyack the girls.
(Henry Fletcher 1908) (Wilkes 1978)
The milk-carters ... sloshed the milk into the cans, chyacked
Dolour about her goggles, and charged out again. (Ruth Park
1948) (Wilkes 1978)
The rowdy bodgie youths kept seats near this group, chiacking
the buxom, brassy-haired waitress as she rushed around with a
tray-load of dishes and lively back-chat. (K.S. Pritchard 1967)
(Wilkes 1978)
'Chiacking', then, is very much a shared entertainment, which both
expresses and promotes the feeling of 'mateship' among those who
jointly engage in it. It is definitely a pleasurable activity, associated with
laughter, rowdiness, noise, and good humour. The following examples
highlight this aspect of 'chiacking':
Pleasant chi-ack in the billets (Action Front, 1940) (AND)
They served out hot tea and in a few moments grumbling gave
place to 'chiacking',. criticism that a few moments ago had been
edged was now good-humoured. (R.H. Knyvett 1918) (AND)
Thus ended the relief of Rustenburg, in cheers and laughter and
chyacking and sleep. (S. Campion 1944) (AND)
They were a vociferous crowd, ruggedly vocal in a loud, chiack-
ing anticipation of the heady joys to come. (E. Lindall 1964)
(AND)
Other types of humour - chyacking and leg-pulling, sardonic
anecdotes, jolliness and exuberance. (Donald Horne 1967)
(AND)
The groomsmen all red in the face and looking as if they would
choke in their stiff white collars, rocked the whole congregation
with a desire to chuckle and chiack. (K.S. Pritchard 1948)
(AND)
168 Speech acts and speech genres across languages and cultures
Though pleasurable for those who engage in it, the activity of 'chiack-
ing' is by no means always pleasurable for those who are the victims of
it. Nonetheless, it is never hostile, and it is expected that it will be borne
with good humour:
Ironbark' s face was red by this time with all the chyacking he got
from the blokes. (D. Stivens 1955) (AND)
Next day at lunchtime I got the same chyacking treatment from
Gordon's brother Frank. (B. Heslin 1963) (AND)
I was always civil to the chaps, for all the chyacking they gave
me. (W.H. Suttor 1887) (AND)
Tommy Bent ... was a victim of most of the 'chyacking'. (Gadfly
1906) (AND)
When their chiacking got too much I would go out and talk to
the turkeys. (M. Eldridge 1984) (AND)
What does it mean, then, to chiack somebody? I propose the following
analysis of the attitude encoded in this concept:
chiack
(a) we want to say something bad about you
(b) because we want to laugh and to feel something good
(c) not because we want you to feel something bad
(d) I think you know: one feels something good if one can do
this with someone like oneself, like a man with a man
Component (a) indicates that chiacking is a collective activity; compo-
nent (b) shows that it is done for pleasure and for fun; component (c)
shows that it is a good-humoured and good-natured activity, devoid of
any hostile intent; and component (d) indicates that chiacking is primar-
ily, though not exclusively, a male activity, either reciprocal (done from
man to man) or collectively (done by a group of men), and that it
implies 'solidarity' and egalitarianism.
The concept of chiacking reflects some of the most characteristic
features of Australian culture: sociability, 'mateship', enjoyment of joint
activities with one's mates (especially idle activities, such as drinking),
male solidarity and male togetherness, associated with displays of 'mas-
culinity' in 'bad language', and so on. The concept of chiacking reflects
also the Australian preference for saying 'bad things' rather than 'good
things', about people in general and about the addressee in particular -
Some Australian speech-act verbs 169
not because one thinks 'bad things' about them or feels 'bad feelings'
towards them, but because of the cultural ideals of roughness, toughness,
anti-sentimentality, anti-emotionality, and so on.
The link between 'saying something bad' and 'feeling something
good' is particularly characteristic. It is the one link which is also
manifested in the typically Australian phenomenon of friendly insults
(G'day ya old bastard!, cf. Taylor 1976), in the tendency to express
enthusiasm by means of swearwords (you bloody beauty!), in the lack
of offensive connotations linked with words such as bugger (poor
bugger), crap, bullshit, and so on.
... the interesting thing about the Australian attitude to human relationship
is the special forms it has to take to avoid coming into conflict with our
basic antipathy towards the public expression of sentiment and emotion.
Because we are unsentimental and cynical towards the emotions, Austra-
lians have to express their social affection in some way which is not on
the face of it self-revealing. Thus, there has evolved the principle of
'rubbishing' your mates and chayacking [emphasis added] the stranger.
In an atmosphere of reciprocal banter or 'rubbishing' Australians can
express mutual affection without running any risk of indecently exposing
states of feeling. (Harris 1962:65-66)
As Renwick (1980:22-23) points out, "with regard to personal charac-
teristics, Australian men and women are friendly, humorous, and
sardonic (derisive, disdainful, and scornful)". They "express negative
feelings and opinions about both situations and people, sometimes about
people they are with". They have a tendency "to be personally evaluative
and to express negative reactions" (1980:29). In particular, negative
remarks play an important role in Australian humour. Renwick observes
that "Americans sometimes feel that Australians' humor is ... disrespect-
ful, harsh, and offensive" (1980:29), and he advises Americans as
follows: "Stand ready, in a relaxed manner, to be tested. The Australian
may challenge you and probe to see if you are a person of substance,
someone with a backbone, some steel inside, some depth and character.
Practice testing and sparring with the Australian yourself." "Develop
personal resilience. Don't be put off by derisive comments, undercutting,
and cynicism." (1980:33-35).
These comments, and this advice, show deep insight into the tradi-
tional Australian ethos - an ethos reflected with particular clarity in the
Australian concept of 'chiacking'. The fact that this key word is now
disappearing from Australian speech, so that the younger generation of
Australians is often unfamiliar with it, reflects some of the changes
170 Speech acts and speech genres across languages and cultures
which this culture is undergoing. As pointed out by many observers of
the Australian scene, since World War II there has been a considerable
shift in Australia from traditional working-class values to middle-class
values. For example, McGregor (1980) reports that according to Gallup
polls taken over the past few decades the proportion of the Australian
population identifying themselves as working class has declined mark-
edly, while the proportion identifying as middle class has correspond-
ingly increased. The decline of the use of key Australian concepts such
as 'chiacking' (and also yarn and shout, to be discussed below) reflects
these broader social changes. In particular, while 'chiacking' is still
a common Australian activity, the concept of 'chiacking' is already
losing some of its salience in the Australian national mentality.
2.2. Yarn
Yarn (which can be used in Australia as either a noun or a verb) is
another important Australian concept, referring to something like a chat
or a talk, but embodying a characteristically Australian way of looking
at the activity in question. It is typically used in the phrase have a yarn;
for example:
They asked the Buxtons to come over to their camp, and have a
'yarn'. (J. Bonwick 1870) (AND)
He used to delight in going to travellers' camps to have a 'yarn'
with them. (M.A. McManus 1913) (AND)
As these examples indicate, 'having a yam' is often seen as a form of
pleasurable sociability. The expectation that 'yarns' generate 'good
feelings' is reflected in the common collocation 'a good yarn', which
implies a satisfying as well as fairly lengthy (and leisurely) verbal
exchange:
You are questioned all about home, what brought you out, and
all such questions, until what is termed in the colony a good yarn
is over, you may then be asked to have a nobler. ('Eye Witness'
1859) (AND)
He says he doesn't really want to do any sort of interview, but
it doesn't take long to see that deep down, the man likes a
good yarn. (Sydney Morning Herald 1986) (AND)
Some Australian speech-act verbs 171
Yarn as a verbal exchange should be distinguished from yarn as a kind of
long tale 'spun' out of facts and fantasy for the purpose of companion-
ship: to have a yarn is not the same as to spin a yarn - another favourite
Australian speech genre of 'the olden days'. But the slow, relaxed nature
of the yarns that people spin (or used to spin) highlights the unhurried,
relaxed nature of the yarn that one can have with someone else.
But although a 'good yam' (with someone) is normally a long one, a
short 'yarn' is also seen as enjoyable, provided that it is leisurely,
unhurried, and without a rigidly imposed temporal boundary. This is
reflected in the common collocation 'a bit of a yam'. For example:
There they all stood and had a bit of a yarn before they
came home. (A.A. Smith 1944) (AND)
The pleasurable, sociable, and unhurried character of 'yarning' is high-
lighted in the following examples:
The manager received me with open arms, and we 'yarned' far
into the night over the old country. (A.W. Stirling 1884) (AND)
I thought it glorious fun smoking our cigars and yarning until
overcome by our long drive, we both fell asleep. (S.S. Junr.
1868) (AND)
But 'yarning' is not an idle activity undertaken solely for pleasure and
devoid of any serious meaning:
By 'yarning', dear reader, I don't mean mere trivial conversa-
tion, but hard, solid talk. (M. Clarke 1896) (AND)
'Yams' differ in this respect from 'chats', which are also exercises in
pleasurable sociability, but which, by definition, play down the signifi-
cance of the exchange. Chatting can be idle, but yarning is not seen as
idle, whatever the topic, because it suggests a serious need for human
contact and for human communication. The following example illus-
trates well this aspect of 'yam':
Some of me old mates from the bush turned up for a beer and a
yarn. (A. Buzo 1986) (AND)
Certainly, 'a beer' and 'a yarn' stand here for enjoyable activities, under-
taken for pleasure. But the sentence illustrates also well the special im-
portance of such activities in the Australian context, where the distances,
the isolation, the loneliness create a special need for human contact,
172 Speech acts and speech genres across languages and cultures
human warmth, and human communication going far beyond the casual,
light-weight sociability characteristic of a chat.
Like chiacking, yarning or having a yarn are concepts with strong
masculine associations. This is another dimension of contrast between a
yarn and a chat; in Australia, men would traditionally have a beer and a
yam (with their 'mates'), whereas 'ladies' would have a 'cuppa' (a cup
of tea) and a chat. These different gender associations may have some-
thing to do with different expectations with regard to 'verbal economy':
the concept of 'chat' implies 'chattiness', that is, a facility with words,
an uninterrupted and easy verbal flow between two people; by contrast,
the concept of 'yam' implies a terseness and a background of silence, of
isolation, and of a real need for a verbal exchange as a form of scarce
human contact, especially with one's friends (one could chat with one's
neighbours every day but one could hardly yarn with them every day).
This need for 'congenial fellowship' (especially male fellowship)
reflected in the concept of yarn is highlighted in this example:
It's hard work sinking bores, and after a few months on your
own, with no one but a couple abos [Aborigines] to yarn to,
you've gotta get stinkin' [drunk] once in a while. (I. Marshall
1962) (AND)
In my English speech act verbs: a semantic dictionary (Wierzbicka
1987) I have posited for chat the following meaning (reproduced here
in a somewhat simplified form):
chat
(a) I want us to say many different things to one another
(b) I think you want the same
(c) I think we will feel something good because of this
(d) I don't think: these things are important
For yarn, I would propose a similar semantic structure but without the
trivialising component (d), and with additional components (a') and (d'),
stressing the participants' need for communication as a form of compan-
ionship with someone like oneself:
yarn
(a) I want us to say many different things to one another
(a') I want to do it for a long time
(b) I think you want the same
(c) I think we will feel something good because of this
Some Australian speech-act verbs 173
(d') I think: it is good if a man can do this from time to time with
someone like himself
Thus the concept of yarn points, indirectly, to the concept of 'mateship',
to the importance of shared activities, to the emphasis on human
relations rather than on productivity or achievement of external goals,
and to the relaxed attitude to time prevailing in Australia.
As pointed out by Renwick (1980), the pace of life in Australia is
relatively slow (at least in comparison with urban/corporate America):
people are less 'task-oriented' and 'future-oriented'; rather, they have a
relaxed, 'day-to-day' orientation, they want to enjoy life and enjoy being
with others, and are more interested in personal relationships (especially
with 'mates ') than in productivity. The characteristically Australian
concept of 'yam' reflects and documents these attitudes. It should be
noted, however, that like the concept of 'chiacking', this concept too
is losing its salience, and that the word yarn is losing ground in
Australian English.
2.3. Shout
Shouting is a specifically Australian concept, standing for an actIvIty
which from early on in the Australian history established itself as
one of the most characteristic national customs, remarked on by all
observers. For example:
Nearly everyone drinks, and the first question on meeting gener-
ally is, 'Are you going to shout?', i.e. stand treat. (W. Burrows
1859) (AND)
'A shout' , in the parlance of the Australian bush, is an authority
or request to the party in waiting in a public-house to supply
the bibulous wants of the companions of the shouter, who of
course bears the expense. (C. Munro 1862) (AND)
Of all the folly that has ever beset a community, that of shouting
has held the ground the longest, and is the most absurd. (Bell's
Life in Sydney 1864) (AND)
He viewed this 'shouting' mania with disgust. (Bulletin (Sydney)
1892) (AND)
174 Speech acts and speech genres across languages and cultures
As many examples from AND clearly indicate, shouting is definitely
linked in Australia with the idea of generosity, and it often is (or was)
asymmetrical, as when a man with money 'shouts' drinks for the
moneyless 'hands' or even for bystanders:
Most peculiar thing to me as the night wore on, and yarn after
yarn went around, the old bloke always shouted, and for all
hands each time. (Western Champion 1894) (AND)
At our approach four miserable derelicts left the stool on the
verandah and slouched into the bar on the prospect of a 'shout'.
(F.l. Brady 1911) (AND)
In the generally non-competitive and super-egalitarian Australian
society, shouting was one domain where one could be freely competi-
tive - competing with other people, as it were, in generosity and in the
spirit of companionship:
In the Westralian mining towns ... man's class is decided by the
number he shouts for ... To shout for the room is common, to
shout for the 'house' nothing extraordinary, and if the shouter
is 'brassed up' at all, he says: 'Call in them chaps outside'.
(Bulletin (Sydney) 1909) (AND)
He was also of that species of good Aussie mixers who, if some-
one 'shouted' a round, would forthwith plonk down a handful
of silver to indicate payment for the next round before anyone
could raise the first glass. (S. Hope 1956) (AND)
At the same time, however, shouting has strong connotations of recip-
rocity and turn-taking:
It is drink, drink, all day, and swim in it at night. Everyone you
meet will 'shout', and you have to 'shout' in return. (Demonax
1873) (AND)
You wouldn't expect a man to leave before his shout would you
Ben? (M. Paice 1978) (AND)
The expected reciprocity of 'shouting' highlights the link which this
concept has with the key Australian value of 'mateship'; and the words
shout and mates frequently occur together:
Some Australian speech-act verbs 175
The unbreakable custom that if four or five mates grouped to-
gether one started to buy all the drinks, but in the circle everyone
had to have his turn. (H.G. Tesher 1977) (AND)
The expectations of reciprocity and turn-taking appear to imply a
mutuality and an equality which is hard to reconcile with the frequent
asymmetry of shouting illustrated earlier. Trying to solve this apparent
paradox I would propose that reciprocity and tum-taking constitute a
social convention associated with shouting but are not a necessary part
of the concept itself. On the other hand, the idea of drinking companion-
ship (male companionship) is part of the concept: even in those cases
when 'shouting' constitutes a one-sided treat and a display of one-sided
generosity the notion is still there that it is good and pleasurable for a
man to drink with other men and that on such occasions it is good to 'do
things' for one's companions and to identify one's own interests with
theirs. Thus, although shouting can be done by one individual, it is
still analogous to chiacking and to yarning in its celebration of relaxed
male companionship, and male solidarity ('mateship'). The following
example illustrates clearly this importance of male companionship and
solidarity over and above any strict reciprocity:
All Merr's mates shouted him at the pub for a week. (A. Garve
1968) (AND)
Thus, the idea of a shout implies not just one invitation to shared
drinking but a sequence of such invitations (typically, a sequence of
'rounds ') and it strongly suggests reciprocity and turn-taking without,
however, precluding one-sided generosity on the part of one particular
person. To account for these facts, I propose to include in the explication
of shouting the following component: 'I think someone will say the
same after me'. It would be natural to interpret 'someone' as 'someone
else', but it can also refer to the speaker himself.
shout
l
(a) I say: I want everyone here to have a drink with me now
(b) I will pay for this
(c) I think we will all feel something good because of this
(d) I think someone will say the same after this
(e) I think we all think: it is good if a man does this with
other men
176 Speech acts and speech genres across languages and cultures
The non-importance of strict reciprocity in the concept of shouting is
highlighted by what AND rightly describes as an extended ('transferred
and figurative') sense of this word (roughly, 'treat offered to someone
else'). For example:
The governor shouted heavy, and gave us all an excellent feed
(N. Earle 1861) (AND)
I'll shout a trip (first-class) for him from Sydney to Narrandera
(Bulletin (Sydney) 1896) (AND)
Once or twice a year I 'shout' the boys of an orphanage to the
pictures. (R. Comm. Moving Picture Industry 1927) (AND)
It's Saturday, and I was wondering if you'd like to have dinner
there. It'll be my shout. It goes on the expense account. (D.
Middlebrook 1975) (AND)
But even this kind of one-sided shout has implications of shared pleas-
ure, as well as of generosity: a person who shouts a treat for someone
else fully expects to share in the target person's enjoyment (if only by
enjoying their enjoyment) and thus shows a generous and friendly
spirit. (The last example above shows a somewhat jarring mixture of
attitudes, and reflects a modem corruption of the pioneer ideal.)
What is truly important about the concept of shouting is the idea of
being generous with other people in the spirit of solidarity and congenial
(male) fellowship. There is no stress on reciprocity in the sense of 're-
payment of a debt' (as in the case of Japanese key concepts of on and
giri, cf. Benedict 1947; Lebra 1974). One is obliged to drink, to share in
the companionship, and to enjoy a relaxed atmosphere of generosity and
group-identification, rather than necessarily 'repay' the treat to the very
person who has provided it. Reciprocity is at the most hinted at by a
general expectation that the recipient would want to do the same (per-
haps some other time, with some other people). Accordingly, 1 have
posited for shout
2
the component 'I think we all think it is good if people
do this with other people', which echoes the last component of shout
l
: 'I
think we all think it is good if a man does this with other men'.
shout
2
(a) 1 say: 1 want to do something good for you
(b) 1 will pay for this
(c) 1 think you will feel something good because of this
(d) 1 think we will all feel something good because of this
Some Australian speech-act verbs 177
(e) I think we all think: it is good if people do this for other
people
It should be noted that this second, extended sense of shout is in fact
growing in use, while the primary sense is declining (together with the
social tradition which gave rise to it, and with the social and cultural
conditions associated with it).
2.4. Doh
If the words chiack, yarn, and shout can be said to affirm and celebrate
'mateship' and 'congenial fellowship' in a positive way, dob in (de-
scribed by OEDS as "Australian slang") can be said to affirm it and
celebrate it as it were in a negative way - by condemning, with
contempt, anyone who betrays it.
AND defines the meaning of the expression dob in as "to inform upon,
to incriminate". But this is not an improvement on the earlier description
offered by OEDS: "to betray, inform against". The notion of 'betraying'
constitutes a crucial difference between the specifically Australian
concept of dobbing and the pan-English concept of informing. On the
other hand, the description 'Australian slang', offered by OEDS, is
misleading: in Australia, dob in is not slang (restricted to some particular
social group), it is simply part of common everyday language, a word
which is in general use and which is clearly one of the key words
in Australian English.
O'Grady (1965) offers the following comment in this connection
(using the word dob in):
Australians are noted for a deep-seated reluctance to report any fellow-
citizen to anyone in a position of authority. Police, bosses, foremen, wives,
etc. must do their own detecting. Anybody who 'dobs in' anybody else is a
'bastard' - in the worst sense of the word. (O'Grady 1965:34)
Similarly, Baker (1959:15-16) mentions "a totally unforgiving attitude
towards 'rats', 'scabs' and betrayers in general" among the most distinc-
tive features of the 'Australian character'. "The essence of the tradition
is loyalty to one's fellows, and the strength of its appeal may be seen in
the restraining power of the term 'scab' in an Australian union."
(Crawford 1970: 137). According to Ward (1958), quoted in Crawford
(1970: 135), "the combination of loyalty to one's fellows with disrespect
178 Speech acts and speech genres across languages and cultures
towards superior orders [and the] enduring disrespect for authority [may
be] traced back to the convicts."
All this is reflected very clearly in the key word dob in. Some
examples:
You said you'd go to the police and dob him in unless he coughed
up. That's the story isn't it? (J. Waten 1957) (Wilkes 1978)
A couple of the Indonesian p.o.w's have dobbed us in. Told the
Nips everything. (R. Braddon 1961) (AND)
In these two examples, dob in could be in principle replaced with inform
on (though not without a significant change in meaning). In the examples
which follow, however, inform on could hardly be used at all, since it
is not used with respect to strictly personal relations (such as, for
example, family relations):
Helen stuck on a real act and dobbed me in to Mum, screaming
about how I had busted her best doll on purpose. (P. Barton
1981) (AND)
Unlike inform on, dob in is derogatory and contemptuous: dobbing is
something a decent person cannot possibly do.
I shut up and let Ray take all the credit. Couldn't dob him in,
could I? (J. O'Grady 1973) (AND)
You bitch! Go and dob me in because I gave you a bit of a shove!
(Williamson 1972:66)
But you feel such a rat to tell on her. To dob her in. (H. F.
Brinsmead 1966) (AND)
The noun dobber is equally, or even more, contemptuous and
derogatory.
Don't look at me, you bastards! I'm no bloody dobber! (J. Pow-
ers 1973) (AND)
The expression 'dobber' was one that I knew implied contempt
and was apt to be applied to tale-bearers and informers. (G.A.W.
Smith 1977) (AND)
One further difference between inform on and dob in is that the latter
implies that the agent is definitely hurting the person spoken of whereas
the former does not necessarily imply that. In informing, the stress is on
Some Australian speech-act verbs 179
the transmISSIon of (potentially damaging) information, not on inter-
personal relations between the speaker and the person spoken of, but
in dobbing in, the stress is on interpersonal relations. This semantic
difference between the two verbs is reflected in a syntactic one. Dob
in treats the victim as a direct object ('to dob someone in') and thus
suggests that the agent is 'doing something to' the person dobbed in. By
contrast, inform on treats the victim as an oblique object (one cannot
'inform someone on'); this suggests that the agent of informing is not
necessarily 'doing something to' the person informed on.
It is interesting to note in this connection that dob, too, can be used
with the particle on, and that dob on is closer semantically to inform on
than dob in is. Inform on, tell on, and dob on all suggest intentional
transmission of damaging information without implying that serious
harm has already been done, as dob in does. At the same time dob on,
which appears to be used mainly by schoolchildren, shares with dob
in its contemptuous and derogatory character: evidently, the general
Australian contempt for those who break group solidarity and who
attempt to side with the authorities against fellow 'subordinates', is an
important part of the Australian school ethos, as well as of the Australian
ethos in general.
I will not try to propose here an explication of dob on, interesting as
it is, focussing instead on the more basic concept dob in, used widely
right across the whole of Australian society.
dob in
I say: person X did something bad
I want you to know this
I think: you will do something bad to X because of this
I know: people like you can do something bad to people like X
and me
I know: X would think that I wouldn't say this to you
I want to say this to you
[people would think something bad about this person because
of this]
[people would feel something bad towards this person because
of this]
It is worth noting that dob in has also another meaning in Australia:
roughly, doing a bad turn to a 'mate' by 'volunteering' for something on
his or her behalf. This meaning is related to the first one insofar as it
implies saying something about a 'mate' to a person in charge, causing
180 Speech acts and speech genres across languages and cultures
something bad to happen to the 'mate,' and thus violating the expecta-
tion of loyalty and mutual support. The main difference between the two
meanings consists in the fact that in one case, one says something bad
about the mate, whereas in the other, one says something unfounded
and embarrassing: namely, that s/he is willing to do something which
in fact s/he is not.
2.5. Whinge
Whinge [wlnd3] (roughly, 'complain' or 'whine') is clearly one of the
key words in Australian English. In other parts of the English-speaking
world it is marginal (although not totally unknown); OED qualifies it
as 'Scottish and dialectal' (noun) and 'Scottish and northern dialects'
(verb), although OEDS hedges this qualification by means of the adverb
,originally' .
The marginal character of whinge and its derivates such as whinger
outside Australian English is reflected by the following examples:
Other local terms for crying ... in Dublin the usual word is
'whinging', hence 'whinger', a term also still used in Cumber-
land, and occasionally heard in Liverpool. (I. and P. Opie 1959)
(OEDS)
Touching the query about 'whinger' ... , 'winjer' was accepted
slang for 'grumbler' at Q. Uni. [Queensland University] a few
years ago, and probably still is. I have seldom heard it else-
where, and no one who uses it seems to know the derivation.
(Bulletin (Sydney) 1934) (OEDS)
The verb whinge, evidently marginal in other varieties of English, in
Australia is a household verb. It plays a crucial role in the socialisation
of children (Stop whingeing!), and in the formation and transmission of
the Australian national ethos. As one observer put it, discussing the
relative unimportance of the value of 'success' and the crucial impor-
tance of the values of 'tough masculinity', gameness, and resilience in
Australian culture:
There is little public glorification of success in Australia. The few heroes
of heroic occasions (other than those of sport) are remembered for their
style rather than for their achievement. The early explorers, Anzac Day:
these commemorate comradeship, gameness, exertion of the Will, suffer-
Some Australian speech-act verbs 181
ing in silence. To be game, not to whinge [emphasis added] - that's the
thing - rather than some dull success coming from organisation and
thought. (Horne 1964:26)
The importance of the concept of 'whingeing' in Australian culture
is reflected in a spectacular way in the very common Australian expres-
sion 'whingeing Poms' or 'whingeing Pommies' - an expression which
shows both the Australian perception of English people and their own
Australian self-image: English people are, above all, 'whingers', whereas
Australians are, above all, 'non-whingers':
The British national pastime of 'grousing' (to use an English
phrase) has given rise in Australia to the derisive expression
wingeing pommy. (Marshall and Drysdale 1962) (AND)
It'll pass a law to give every single wingein bloody Pommie his
fare home to England. Back to the smoke and the sun shining ten
days a year and shit in the streets. Yer can have it. (T. Keneally
1972) (AND)
Whingeing Poms make me ill. (W.F. Mandle 1974) (AND)
What exactly is whingeing? Clearly, it is a concept closely related to
complaining. But, first, complain is neutral, and does not imply any
evaluation of the activity in question, whereas whinge is critical and
derogatory. Furthermore, complain is purely verbal, whereas whinge
suggests something that sounds like an inarticulate animal cry. Being
purely verbal, complaining can be seen as fully intentional, whereas
whingeing can be seen as only semi-intentional and semi-controlled.
Finally, whingeing, like nagging, and unlike complaining, suggests
monotonous repetition.
In English speech act verbs (Wierzbicka 1987) I posited for complain
the following semantic structure (reproduced here in a slightly
simplified form):
(a) I say: something bad is happening to me
(b) I feel something bad because of this
(c) I want someone to know about this
Whinge appears to attribute to the speaker (the whinger) an analogous,
but more elaborated, attitude:
whinge
(a) I say: something bad is happening to me
182 Speech acts and speech genres across languages and cultures
(b) I feel something bad because of this
(b') I can't do anything because of this
(c) I want someone to know this
(d) I want someone to do something because of this
(e) I think no one wants to do anything because of this
(f) I want to say this many times because of this
Component (b') of this formula suggests a feeling of total baby-like
helplessness, (d) indicates passivity and reliance on others, (e) suggests
an element of infantile resentment and self-pity, whereas (f) spells out
the reliance on the equally infantile 'strategy' of monotonous repetition
(as in an infant's crying). Generally, then, the concept of whingeing
likens the attitude of those who indulge in it with that of crying babies,
and what Australians think of people who behave like crying babies is
best expressed in another important Australianism: the noun sook [srok]
(adj. sooky). Some examples:
(He goes to her and holds her gently ... She sobs a little, but then
forces a laugh and leaves him.) Ruby: Well! You'll think I'm a
sook. (R.I. Merritt 1975) (AND)
Annie felt sick with fear. 'Sookie sook, I'm going to tell on you' ,
chanted Rosa. (Australian Short Stories 1985) (AND)
The girl applied a hefty hip ... and flattened him. Sprawled on
the bitumen, he began to howl. 'Bloody sook!' said the girl,
disgustedly. (Bulletin 1986) (AND)
As pointed out by Horne (1964:40) among others, Australians are
cheerful and practical-minded optimists. They admire toughness, resil-
ience, good humour, and rough 'masculinity' - in hard times as well as
in good times. Their folk heroes are Ned Kelly and various other real
or legendary 'wild colonial boys', for whom the important thing was not
so much to live in comfort and security or to succeed as to:
... die hard, die game, .
die fighting, like that wild colonial boy,
Jack Dowling, says the ballad, was his name.
(a poem by John Manifold, quoted in Ward 1958:217)
Acccording to the same Australian ballad, "'I'll die but not surrender',
said the Wild Colonial Boy" (Wannan 1963:17).
The present generation of Australians think, it seems, less of 'dying
hard' and more of 'having a good time' (on the growing hedonism of
Some examples of complex speech genres 183
Australians see Conway 1971; King 1978); but the contempt for
'sooks' and 'whingers' has remained part and parcel of the present-day
Australian ethos. Australian national mottos are still 'no worries' and
'she'll be right'. Australians still admire actions rather than words or
ideas. They value practicality and self-reliance. They also assume and
approve of mutual reliance of 'mates' on one another (cf. Renwick
1980:16). But any 'babyish' reliance on more powerful 'other people'
and 'babyish' indulgence in repeated 'crying' (instead of a search for
practical solutions) is totally incompatible with the Australian 'bush
ethos' and with whatever remains of it in the modern Australian
mentality. The key verb whinge reflects and documents these attitudes.
A full study of Australian colloquial speech-act verbs would have to
include many more, such as stir, sledge, skite, rouse on, pimp on, ear-
bash, big-note oneself, knock, (w)rap (up), and fang (see Wilkes 1978;
AND). I believe, however, that the five that have been discussed here,
are particularly representative, and particularly important.
3. Some examples of complex speech genres
3.1. The black English dozens
I will start with a genre that is well known, especially through Labov
(1972): Black English 'ritual insults'. As Labov and others have pointed
out, the genre in question has a number of different folk names, which
reflect some regional and, apparently, semantic variation, sounding and
playing the dozens being perhaps the most common among them. It is a
form of 'street talk', engaged in by black adolescent boys, a kind of
verbal contest. As Abrahams (1974:241) points out, speaking of black
culture: "Playing ... is an important way in which one distinguishes
oneself in public, and engaging in witty verbal exchanges is one impor-
tant way of playing ... Active verbal performance in the street is one
of the main means of asserting one's presence and place." The wit and
the verbal virtuosity are exercised mainly through breaking the rules
and conventions of the 'respectable' society.
The cultural significance of this genre is excellently captured in a
passage from Mezzrow and Wolfe, quoted in Abrahams (1974:240):
184 Speech acts and speech genres across languages and cultures
These boys I ran with at The Corner, breathing half-comic prayers at the
Tree of Hope, they were the new sophisticates of the race, the jivers, the
sweet-talkers, the jawblockers. They spouted at each other like soldiers
sharpening their bayonets - what they were sharpening, in all this verbal
horseplay, was their wits, the only weapons they had. Their sophistication
didn't come out of moldy books and dicty colleges. It came from opening
their eyes wide and gunning the world hard ... They were the genius of
the people, always on their toes, never missing a trick, asking no favours
and taking no guff, not looking for trouble but solid ready for it. Spawned
in a social vacuum and hung up in mid-air, they were beginning to build
their own culture. Their language was a declaration of independence.
(Mezzrow - Wolfe 1969:193-194)
It is not my purpose to try to add anything to the understanding of the
ge:lre in question, of which I have no first-hand knowledge. All I would
like to do is to propose a semantic formula, constructed in the natural
semantic metalanguage, which spells out the illocutionary force of this
genre, as I understand it on the basis of studies such as Labov (1972) and
Abrahams (1974). It seems to me that a succinct formula of this kind
'sums up' the genre in question in a metalanguage which is essentially
culture-independent and thus facilitates cross-cultural comparisons.
One characteristic (though uncharacteristically mild) example may be
in order: "Your mother so old she got spider webs under her arms"
(Labov 1972:312). The semantic formula:
(a) I want to say something bad (about your mother)
(b) I think: everyone here knows: I don't think this (about your
mother)
(c) I want to say something that some people would say is bad
(d) I say this because I want people here to feel something good
(e) and to think something good about me
(f) I want people here to think that I can say things that other
people can't
(g) I think after this you will say something like this (about my
mother)
(h) I think you will want to say something more bad than this
(i) I want you to say it if you can
(j) I think we can say things like this to one another because we
are becoming men
(k) I think we will all feel something good because of this
Some examples of complex speech genres 185
Essentially, the speaker is saying something 'bad' and obscene about
the addressee's mother (component a), something that is clearly not
true (component b), trying to be daring (c) as well as ingenious (f),
inviting the addressee to respond in kind (g) and challenging him to
outdo him (h and i). In doing so, he is trying to entertain the non-
addressed participants (d) and to gain their admiration (e), as well as
participate in the shared 'fun' himself (k); he is also trying to assert and
validate the group's view of themselves as approaching manhood U).
It is interesting to note that certain aspects of this genre as explicated
here correspond quite closely to certain features of the black English
ethnography of speaking, discussed earlier; in particular, I have in mind
the positive attitudes to 'boasts' and the desire to entertain the audience
(cf. Chapter 3 above).
It is also interesting to note some links between the Black English
'dozen' and the Australian English 'chiacking': both are collective male
activities undertaken for fun, involving an audience, and consisting in
saying offensive things about the addressee, without the intention to
offend. But the differences between the two genres are as striking as the
similarities. In particular, the 'dozen' stresses verbal virtuosity, whereas
the Australian ethos values grunts and discourages eloquence; the
'dozen' is competitive, whereas the 'old chiack' is cooperative (with
'mates' acting as a group rather than as separate and competing
individuals); the 'dozen' is obscene and reaffirms the group's rejection
of the values and taboos of the society at large, whereas the 'chiack' is
not obscene, and reaffirms the prevailing values of the traditional
Australian society (while emphasising the distinct character of this
society, defined partly by its rejection of, and opposition to, 'Pommy
ways').
3.2. The Hebrew 'dugri talk'
According to Katriel (1986), a 'dugri talk' is a kind of speech event
which plays a crucial role in social interaction in the contemporary
Israeli society. It is a kind of frequently enacted social ritual which takes
to its logical conclusion the more general 'dugri' mode of speaking,
highly valued in the Israeli 'Sabra culture'. Dugri, a loan from Arabic,
means, literally, 'straight', and sabra is the name of a native Israeli fruit,
a kind of prickly pear. As Oring (1981:24, quoted in Katriel 1986:19)
points out, "the sabra fruit is a metaphor for the native personality. Like
186 Speech acts and speech genres across languages and cultures
the prickly pear, the native born is sweet and gentle within, but only to
those who understand how to penetrate the tough and thorny exterior."
This 'tough and thorny exterior' of typical Sabras manifests itself,
above all, in their tendency to 'talk straight (i.e. dugri)'. The sentence
"He is dugri" means in Hebrew "that the speaker tends to be direct and
straightforward in expressing his non-complimentary thoughts or
opinions" (Katriel 1986: 15).
There are good historical reasons for the great value attached in
Israeli culture to the 'dugri' mode of speech, explored in an illuminating
way in Katriel's book (and also in Oring 1981). Katriel (1986:21-22,31)
mentions in particular: "the assertiveness cluster, which has been
associated with the revolutionary orientation of the Zionist movement
encapsulated in the phrase 'the negation of the Diaspora"'; "the Sabra
culture's emphasis on simple, manual agricultural labor as a means of
getting away from the Diaspora image of the Jew as a luftmensch"; the
"rejection of decadent European ways of speaking that involve twisting
the forms of speech for the purpose of showing respect"; the substitution
of solidarity, cameraderie and a spirit of 'communitas' for respect,
hierarchy, and distance-creating courtesy; and so on.
According to Katriel, all of the cultural values associated with the
Sabra culture (a cult of solidarity, simplicity, sincerity, frankness,
'directness', 'straightforwardness', 'truthfulness', 'assertiveness', and so
on) find their best expression in the 'dugri' mode of speech, and are
epitomised, in particular, by the 'dugri ritual', that is, by the 'dugri talk'.
In native terms, this event is referred to as siha dugrit, a dugri talk. A dugri
talk is not just any encounter in which the dugri idiom is employed or in
which utterances indexed as dugri are exchanged. A dugri talk is a distinct
speech event with a sequential and motivational structure of its own.
(Katriel 1986:57).
Katriel offers, among others, this example: "An engineer in his early
thirties told me at some length about a dugri talk he initiated with his
boss. He started what he described as siha dugrit by declaring: 'I want
to speak to you dugri. I don't like the way this department is being
run. ,,, Describing a typical example of this genre, Katriel writes:
It was a ritual act of confrontation, a ceremony of discord, performed in the
culture's legitimising idiom: the idiom in which one's integrity and one's
shared cultural world are reaffirmed. The use of dugri speech here, as in all
other cases of its ritual enactment, served to counteract what in the Sabra
culture is considered the tendency to gloss over interpersonal differences
Some examples of complex speech genres 187
in the service of a false, superficial consensus, a concern with harmony in
interpersonal relations at the expense of dealing with basic issues and
matters of principle. Despite the discomfort caused by the confrontational
tone, the dugri ritual was experienced as a moment of true contact, of
unmasking, and was received as both legitimate and appropriate even by
participants whose own style was a far cry from dugri speech. (Katriel
1986:58-59)
This general characterisation is very helpful, but it cannot replace a
rigorous definition, formulated in a language which would facilitate
cross-linguistic and cross-cultural comparisons. I would propose the
following:
a dugri talk
I think something bad about you
I want to say this to you
I know: you don't think the same
I know:
if I say this to you you can feel something bad because of this
I know: someone can think:
I will not say it to you because of this
I don't want not to say it because of this
I think:
it would be bad if I didn't say it to you
you are someone like me
you and I can say things like this to one another
because you and I want the same kind of thing
people think:
it is good if people can say to one another what they think
It is particularly interesting to compare the ethos reflected in the 'dugri
talk' (as represented in this formula) with the Australian ethos, reflected
in chiacking, rubbishing, or in friendly insults (G'day ya old bastard).
Both Israeli culture and Australian culture can be described, and have
been described, as cultures which value 'directness', 'straightforward-
ness', 'terseness', 'simplicity of speech', 'plain talk', and so on; and
as cultures whIch value 'solidarity', 'equality', 'cameraderie', 'coopera-
tion', and so on; as cultures which dislike 'artificial politeness', 'social
graces', 'glib (or smooth) talk', 'polite polysyllables', and so on, as
cultures which encourage people to express negative reactions, and to
say 'unpleasant things' rather than 'pleasant things' to one's addressee.
188 Speech acts and speech genres across languages and cultures
All this is true (in some sense) and yet global descriptions of this kind
conceal profound differences which are as real as the similarities. In
particular, Australian ethos does not encourage people to express freely
their 'bad thoughts' about the addressee. For example, it is one thing to
call someone, to their face, 'an old bastard' (making clear, at the same
time that one doesn't think this), and another, to tell them that one really
thinks bad things about them. Israeli culture encourages the speaker to
reveal to other people 'bad thoughts' that he or she has about them;
Australian ethos does not encourage that. In Australia, people often
say 'bad things' when they think or feel 'good things'. But this is not
the kind of 'directness' celebrated in Israel as 'dugri speech'.
Australians, too, could be described, with some justifications, as
'prickly pears'. But metaphors, like global labels, are often misleading.
Explications couched in semantic primitives force us to be precise, and
enable us to identify the real similarities as well as the real differences.
3.3. The Polish kawal
There are few speech genres as important in Polish culture as kawaf
(plural kawaly). Roughly speaking, kawat is a kind of joke. But Polish
lexically recognises other kinds of jokes: dowcip and zart. tart is not
necessarily verbal or not only verbal; it may correspond to what is called
in English a practical joke. Dowcip is necessarily verbal, and so is
kawal. But kawal differs significantly from dowcip in its repeatability
and in its ingroupness. Dowcip as a mass noun means 'wit'; a dowcip (as
a speech genre) can be used as the nearest translation equivalent of
'witticism': it evokes the idea of verbal creativity. Of course, people can
repeat old dowcipy (PL) to one another, but the word itself evokes an
original display of wit by some creative individual. A kawai, on the other
hand, contains no reference to individual wit. It is conceived of as an
anonymous creation of oral culture, as a cultural coin which is meant for
general circulation. Every kawal expresses a bit of collective wisdom,
collective experience, collective outlook. It promulgates ingroupness,
solidarity, social integration vis-a.-vis some outsiders. Thus, the
prototypical kawaly (PL) are political. They express national solidarity
vis-a.-vis foreign powers: the Nazi occupation during World War II, the
Soviet-imposed communist regime in post-war Poland, the foreign
partitioning powers in the nineteenth century.
Some examples of complex speech genres 189
In addition to political kawaly, which have played a colossal role in
Polish culture, there are also kawaly which correspond to English dirty
jokes. These usually express male solidarity and male ingroupness.
Thus, kawal is a folk genre stressing ingroupness and wide circula-
tion. One enjoys hearing a kawal not only because of its humourous
value but also because of the feeling of belonging it gives one. In contra-
distinction to dowcipy, kawaly are not valued for their ingenuity or
sophistication. Etymologically, kawal is an augmentative (from kawalek
'a piece'), and this augmentative character is still felt, carrying the
implication that a kawal has nothing over-sophisticated about it, that it
is something 'rough' and 'thick', something that can be widely shared.
(A kawal chleba 'a piece-AUG of bread', is a very thick and inelegant
piece of bread, but the connotations are positive, not negative; the
expression implies a hungry man's point of view.) Thus, the ingroupness
of a kawal has nothing elitist about it. The group to which a kawal refers
is seen as a very broad and strong one. Nonetheless, a kawal also has a
certain conspiratorial quality: the group within which a given kawal can
circulate wants to exclude outsiders.
I hypothesise that the semantic structure of the word kawal contains
the following component: 'I think I can say this to you because we think
the same about things of this kind'. The implication is: I can tell you, but
there are people to whom I couldn't tell it. This conspiratorial air of
kawaly is often felt to be the saving grace of a joke which would be
rather poor if it were assessed in terms of its sophistication or elegance.
As a genre, kawal is firmly rooted in Polish history, in the specific
conditions of Polish life in the nineteenth and twentieth centuries. It is
an expression of a counter-culture, subsisting in defiance to the official
culture imposed from outside. It reflects the persistent need of the
nation for the therapeutic effect of a shared laugh. The presumed wide
circulation of kawaiy written into the very structure of the concept
kawal fulfils an important social function as an expression of widely
felt solidarity vis-a.-vis some powerful outsiders, as a way of keeping
up the spirit of defiance, as a means of psychological self-defence of
the nation.
This psychological role of kawaiy is clearly reflected in the way
they have tended to correlate with political change in the country - not
only in their content but also in their relative frequency at any given
time. For example, as pointed out by Garton Ash (1983) and other
observers of the Polish scene, the period of national euphoria due to the
emergence of Solidarity in 1980-81 witnessed a marked decrease in the
190 Speech acts and speech genres across languages and cultures
production and circulation of kawaly, whereas the bleak time after the
imposition of martial law in December 1981 saw their renewed growth.
Since August [1980] the consumption of political jokes, like the consump-
tion of alcohol, had dwindled - people had other, better outlets for their
political energy and ingenuity ... (Garton Ash 1983: 106)
The incredible claims of official propaganda were swept away in a torrent
of popular jokes - the acid political humour which returned with martial
law, for this kind of humour is an expression of impotence as much as
defiance. Every new demonstration, strike or protest was ascribed by the
TV news to tiny groups of Solidarity 'extremists'. The Poles translated this
with the 'TV Dictionary':
2 Poles: an illegal gathering
3 Poles: an illegal demonstration
10 million Poles: a handful of extremists.
(Garton Ash 1983:271-272)
Normally, a kawal requires some kind of introduction: addressees
must realise in advance that what they are going to hear is a kawal.
Typically, this introduction comes in the form of a question: 'Do you
know this kawal?', (the assumption being that the kawaly circulate
widely and are likely to be known to the addressee). By contrast, jokes
can come unannounced (for example, in the course of a lecture, a talk, or
a serious conversation). Kawaly (like parties, parlor games, or social
drinking) are meant to promote pleasant togetherness, that is, they are
meant to make the addressee and the speaker feel good together. I
think jokes have that function too. Witticisms, on the other hand, don't:
they seek the addressee's admiration for the speaker rather than
shared pleasure.
But even though jokes, like kawaly, are meant to promote a pleasant
togetherness, they do not have to express and promote group solidarity.
For example, imaginary book titles, such as 'All about dogs, by K. Nine',
'Say your prayers, by Neil Down' or 'The world of vegetables, by
R.T. Choke', can be listed in a collection of jokes, but their Polish
equivalents could not be listed in a collection of kawaly. The reason is,
I think, that while such imaginary titles are funny, they do not appeal
to any particular attitudes shared by some particular social group. But
kawaly always appeal to such shared attitudes.
Finally, a kawal has a point, which is not expressed overtly and which
has to be grasped ('got') by the addressee. Typically, jokes, too, have a
point to be discovered by the addressee ('Did you get it? I don't get it. ').
Some examples of complex speech genres 191
But in the case of a joke, this doesn't seem to be absolutely necessary.
If an adult says to a plump little girl, "I like you a lot, I'll eat you with
plum sauce", it is a joke, but one without a point to be discovered,
beyond the mere fact that it wasn't really meant and that it was said
to cause the little girl to laugh and to feel good. On the other hand, a
kawai, with its fictitious little script, always poses a mental task for
the addressee. One can joke with a baby, but one can hardly tell a baby
a kawal.
Of course, there is a difference between joking and telling jokes. One
wouldn't tell a baby a joke either. But it seems that when the noun joke
is used, the differences between joking and telling jokes are disregarded,
and the two concepts are subsumed under one. This broad concept,
identified by the noun joke, has no Polish equivalent, presumably due to
the importance of the specialised concept kawal and the concomitant
restriction of the remaining members of the same semantic field.
We can now attempt semantic formulae for kawal and joke.
kawal
(a) I want to say something to you of the kind that many people
say to one another
(b) I say: one can know this (X)
(c) I think you know that this is not true (Le. that one can't know
this)
(d) I say this because I want you to laugh
(e) I think you know that when I say it I want you to think
something that I don't say
(f) I think when you think of it you will laugh
(g) I think we will both feel something good because of this
(h) I think I can say it to you
because you and I think the same about things of this kind
and feel the same when we think about them
Component (a) of this explication shows that kawa/y have to circulate
widely; components (b), (c), (d) and (g) are shared by jokes, and jointly
spell out the fictitious character of what is being said, the intended
humour, and the intended shared pleasure; component (e) indicates that
there is a 'point' that the addressee has to 'get'; and (f) shows that it
is this 'point', which has to be reconstructed by the addressee, that is
expected to be a source of laughter; component (h) spells out the
in-groupness of kawaly and the assumption of shared attitudes.
192 Speech acts and speech genres across languages and cultures
joke
(a) (-)
(b) I say: one can know this (X)
(c) I think you know that this is not true (i.e. one can't know
this)
(d) I say this because I want you to laugh
(g) I think we will both feel something good because of this
3.4. The Polish podanie
Writing in 1983, the British historian Timothy Garton Ash characterised
the socio-political situation in post-war Poland in the following words:
This regime can accurately be described as 'totalitarian' in the sense that it
aspires to total control over every aspect of its citizens' lives, to break
every social bond outside its aegis, to destroy what the Enlightenment
philosophes called 'civil society'. (Garton Ash 1983:8)
Other western observers of life in communist Poland have made similar
remarks. The key words and phrases constantly recurring in this litera-
ture include the following: 'total control' (Garton Ash 1983:8), 'bureau-
cratic controls' (Davies 1981, 2:597), 'the bureaucratic Leviathan'
(Hirszowicz 1980), 'petty bureaucrats' (Davies 1981, 2:617), 'party-state
bureaucracy' (Kolankiewicz - Lewis 1988:24), 'official lawlessness'
and 'petty despots' (Davies 1984:41), and so on.
Writing more generally about the Soviet bloc, Davies commented:
In the Soviet Bloc, there is no higher authority than the Kremlin. There is
no rule of Law above the dictator of the day.... What is more, in relation
to the mortals beneath him, his particular mode of the 'dictatorship of
the proletariat' can be imitated by all the descending hierarchy of petty
despots right down the endless links of the political chain. From the
supreme web of the Soviet nomenklatura with the Grand Spider at its
centre, web upon tangled web radiates out from the Kremlin into the
farthest reaches of the Soviet empire ... Every web has its 'spider'; and
the spiders, even the benevolent ones, recognise no Law higher than
their own. Such a state of official lawlessness is so alien to Americans and
to West Europeans (not to mention the Japanese) that most serious
attempts to describe it are instinctively dismissed as flights of fancy.
(Davies 1984:40-41).
Some examples of complex speech genres 193
Speaking more specifically about communist Poland, Davies pointed
to the profound division between the 'power' (w/adza) and 'society'
(spo/eczenstwo), between the bosses and the people; and he commented:
... it is undeniable that what Milovan Djilas named the 'New Class', and
what others have called 'The Bureaucratic Leviathan', constitutes the most
characteristic feature of the supposedly classless societies of Eastern
Europe. . .. In Poland, where the regime has virtually no legitimacy, it is
the source of common oppression. (Davies 1984:45).
The socio-political realities of life in communist Poland found innu-
merable reflections in the Polish language (cf. for example Wierzbicka
1990). In the present chapter, I will point to just one such reflection,
particularly relevant to the area of speech acts and speech genres: to
the central importance of the concept of podanie in everyday life in
communist Poland.
The monumental Dictionary of the Polish Language (SIP 1958-68)
defines the word podanie as follows: "pismo skierowane do wladz z
0 cos; petycja", 'a written document addressed to the authorities,
asking for something; a petition'; and Skorupka' s (1974) phraseological
dictionary of Polish offers a shorter version of the same: "pismo z
prosbfl 0 co", 'a written document asking for something'. Podanie,
then, is a special genre, involving communication between 'people' and
'authorities'; and this communication consists, invariably, in 'people'
asking' authorities' for 'favours' and presenting themselves as dependent
on the authorities' good will.
Clearly, the word podanie has no equivalent in English - not because
in English ordinary people never ask any authorities for 'favours', but
because the idea that individuals have to ask authorities for 'favours' is
not sufficiently salient in English-speaking countries to have led to the
emergence of a special concept, and of a special speech genre. But in
Polish, the idea is (or was) very salient, no doubt because ordinary
people's lives in communist Poland were dominated, to a considerable
degree, by their dependence on the arbitrary decisions of bureaucratic
'despots'. An interesting item of linguistic evidence for this is furnished
by the set phrase papier podaniowy, 'paper for writing podania (PL)' -
the usual way of referring to A4 writing paper (a phrase given a special
entry in both SIP and in Skorupka 1974).
Skorupka's (1974) phraseological dictionary offers the following
common collocations involving podanie: "podanie do ministra, do
dyrektora; do dziaru kadr, do 'a podanie to the minister, to the
194 Speech acts and speech genres across languages and cultures
director; to the Personnel Department, to the court'; "podanie 0 co:
podanie 0 stypendium, 0 na 'a podanie
(asking) for a scholarship, for admission to a university'; "odrzucic,
podanie", 'to reject, to accept a podanie'; "zalatwic podanie
(odmownie, pozytywnie)", 'to settle a podanie (by refusal, or
positively)' .
Thus, a podanie is a written communication from an individual to an
institution, asking for something that the institution mayor may not
grant, where the response is seen by the petent ('supplicant') as arbitrary
and unpredictable, and yet as indispensable to normal conduct of life.
Hardly any aspect of people's lives in communist Poland, no matter
how trivial, could be conducted without the need to write podania - and
to wait for the response, hoping that it might be benevolent. For ex-
ample, a university student asking for an extension of the deadline for
submitting a thesis, or an employee asking for permission to take annual
leave at a particular time, submitted a podanie. In similar circumstances
in an Anglo-Saxon society, it is often sufficient to write a letter, that
is, something that is seen as belonging to the same genre as private
letters. The closest Polish equivalent of letter is list, but in communist
Poland a student or an employee would never seek official favors or
concessions by means of a list. The decision of the institution would not
be communicated to the requester in the form of a list either; what that
person c<?uld hope to receive as a reply was not a list but a pismo, that
is, a bureaucratic document categorised as something quite distinct
from personal letters. There is (or was) a big difference between, say,
a list od dziekana 'a (personal) letter from the Dean' and a pismo z
dziekanatu 'an official communication from the Dean's office'.
Of course English, too, has special words for formal, bureaucratic
speech genres, addressed by an individual to an institution, such as appli-
cation. But application is restricted to situations when an institution has
the power, but also the obligation, of offering certain categories of people
certain rights and goods if these people meet certain specified criteria.
Thus, one writes an application (and not a letter) seeking to obtain
certain positions, scholarships, and so forth, which have been publicly
advertised, with the relevant criteria explicitly formulated. So while it is
up to the addressees to decide whether a given applicant meets the re-
quirements, the applicant is not seeking a kind of personal favor, like
a person writing a podanie. The concept of application presupposes a
certain situation, a situation which can often be presented in
a standardised application form (often one doesn't even have to write
Some examples of complex speech genres 195
an application, it is sufficient to fill out the form). The assumption is
that the addressees will be guided by explicit criteria and that they have
no 'arbitrary power over individuals.
Typically, a podanie starts with a phrase such as Uprzejmie prosze
... or Niniejszym zwracam ~ z uprzejma prosbfl ... 'I ask politely',
'hereby I address you with a polite request (for a favor)'. But it would
be odd to couch an application in some such terms, because a person
who is writing an application is applying, not asking for or requesting.
(For the difference between ask for and request see Wierzbicka 1987.
The Polish words prosic (verb) and prosba (noun) are in fact closer to
ask for than to request.)
It is also worth noting that while an application may be 'unsuccessful'
it can hardly be 'refused' or 'rejected'. Furthermore, the author of an
unsuccessful application may be even thanked by the addressee for
applying ('Thank you for your application ... '). But it is unthinkable that
the bureaucrat to whom a podanie is addresseed could 'thank' the writer.
Similarly, applications can be, and often are, 'invited' (for example
'applications from suitably qualified persons are invited ... '); but it is
unthinkable that podania could be 'invited'. The reason is that people
handling applications can be seen as wanting people to apply, but in
the case of a podanie, the writer is (or was) seen as a 'petitioner', a
'supplicant', not as someone who can be seen as cooperating with the
addressee.
This discussion leads us to the following explications:
podanie
I say: I want something (X) to happen to me
I know: it cannot happen if you don't say you want it to happen
I say this because I want you to say you want it to happen
I don't know if you will do it
I know many people say things like this to you
I know you don't have to do things that people want you to do
application
I say: I want something (X) to happen to me
I know: it cannot happen if you don't say you want it to happen
I think many people may want the same
I think you want some people to say to you that they want it
I say this because I want you to say you want it to happen to me
I know you cannot do it if I don't say some things about me
196 Speech acts and speech genres across languages and cultures
I say these things here because of this
I think you will say if you want it to happen to me
Of course, every system can be abused, but the language-specific
taxonomy of speech genres reflects what is seen as the typical situation.
Contrasting concepts such as application and podanie reflect what native
speakers perceive as forms of life typical of their society.
4. Conclusion
Nearly three centuries ago, John Locke wrote:
A moderate skill in different languages will easily satisfy one of the truth
of this, it being so obvious to observe great store of words in one language
which have not any that answer them in another. Which plainly shows that
those of one country, by their customs and manner of life, have found
occasion to make several complex ideas, and given names to them, which
others never collected into specific ideas ... Nay, if we look a little more
nearly into this matter, and exactly compare different languages, we shall
find that, though they have words which in translations and dictionaries are
supposed to answer one another, yet there is scarce one of ten amongst the
names of complex ideas ... that stands for the same precise idea which the
word does that in dictionaries it is rendered by. (Locke 1959 [1690], 2:48)
I believe that few areas of language illustrate the truth of Locke's
statement better and more revealingly than the area of speech acts and
speech genres. Language-specific terms for speech acts and speech
genres present linguistically codified modes of social interaction. Precise
semantic analysis of such terms should, it seems to me, provide a new
source of insight and rigour in cross-cultural studies.
Chapter 6
The semantics of illocutionary forces
1. Are illocotionary forces indeterminate?
Common sense suggests that human life consists, to a very large extent,
of a variety of speech acts. From morning to night, people ask, answer,
quarrel, argue, promise, boast, scold, complain, nag, praise, thank, con-
fide, reproach, hint - and so on and so on. Moreover, from morning to
night, they seek to interpret (consciously or subconsciously) what other
people are doing when they speak, that is, what kinds of speech acts
they are performing. Virtually every time someone opens his or her
mouth in our presence we need to interpret their utterance as this or
that kind of speech act. Was it a threat? Or just a warning? Was this a
suggestion or a request? Was this a criticism or just a casual remark?
Was this a hint?
The idea that in speaking people perform different kinds of acts, and
that the semantic and/or syntactic structure of an utterance may depend
on the kind of act being performed, is very old. It goes back at least as
far as the Stoics.
4
In modern times, too, similar ideas have been put
forward by a number of different thinkers - notably, by Josef Schachter
(1935) in his Prolegomena to a critical grammar, by Ludwig Wittgen-
stein (1953) in his Philosophical investigations, and by Mixail Baxtin
(1952) in his Speech genres. But it was of course J.L. Austin (1962)
who expressed and developed this idea in a way which attracted the
attention of present-day linguists, and through whose work it was intro-
duced into modern linguistic theory.
As a result of Austin's influence, a new model of speech emerged in
linguistics, according to which whenever people speak they produce
sequences of identifiable speech acts: they greet, they ask, they invite,
they suggest, they warn, they apologise, and so on. In fact, it was sug-
gested (by 'generative semanticists', for example Ross 1970) that every
sentence contains in its semantic structure (which is also its underlying
syntactic structure), a clause that identifies the nature of the speech act
performed by means of the sentence. For example, a sentence such as:
198 The semantics of illocutionary forces
What time is it? was said to be derived from I ask you what time it is,
and a sentence such as Come here! was said to be derived from I order
you to come here.
Actually, this 'new' model of analysis, too, was a reinvention of a
mode of analysis developed many centuries earlier - notably, by Peter
Abelard in the twelfth century, by Roger Bacon in the thirteenth
century, or by Paul of Venice in the fourteenth century.
5
The close
parallels between the speech act analysis advanced by linguists in the
second half of the twentieth century, and that developed by their
mediaeval predecessors, are quite fascinating, and it is strange that they
are never mentioned in contemporary linguistic writings on speech
acts. This is not the place, however, to comment any further on these
parallels, or on the historical lessons which could be drawn from them.
But contemporary syntactic theories tend to have a rather limited life
span. After a few years of vigorous activity, 'generative semantics' lost
confidence in itself, and disintegrated, and the newly reinvented idea of
deriving illocutionary forces from underlying illocutionary verbs lost
most of its adherents.
Nonetheless the view that the illocutionary force of an utterance is a
part of its semantic structure has survived - at least to the extent of
being still regarded as a worthwhile target for attacks. Such attacks are
increasingly frequent and increasingly vigorous (cf. for example Bach -
Harnish 1982; Leech 1983). The attackers point out that most utterances
can't be identified, unambiguously, as instances of a particular speech
act. For example, Come here! can be an order but it can also be a request.
This gun is loaded can be a warning, but it can also be a statement of
fact - or a threat. How are you? can be a question but it can also be
a greeting. And so on.
The so-called performative analysis forced the analyst to reconstruct
in each case one explicit performative verb, but in fact - it is now
claimed - most utterances are indeterminate. In other words, when we
hear people speak, we may understand perfectly well what they are
saying but we usually don't know what exactly they are doing: warning?
threatening? boasting? suggesting? promising? We can of course make
guesses as to the speaker's illocutionary intentions, but we can never be
sure if these guesses are correct. Sometimes the situation and the context
can make it fairly clear what the speaker is doing (for example, that he
is threatening rather than warning), but even then (it is claimed) our
interpretation is guided by 'pragmatic considerations' rather than by
linguistic clues.
Are illocutionary forces indeterminate? 199
And what applies to ordinary speakers and listeners applies also to
linguists. As linguists - so the argument goes - we cannot identify
the illocutionary force of any utterances that we may wish to analyse.
Consequently, illocutionary forces are outside the province of linguis-
tics; they are a concern of pragmatics, but not of syntax or semantics.
For example, a linguist qua linguist cannot distinguish between
expressions such as Go to London! and expressions such as Go to hell!
(both 'commands'), or between How are you? and How old are you?
(both 'questions').
1 disagree with this view, and 1 believe that it misrepresents the nature
of human communication. I believe that when we listen to other people
we more often than not do know what they are doing, and we know it,
to a large extent, due to unmistakable linguistic clues. Intonation no
doubt plays an important role in this respect (see below, section 7.6),
but even leaving intonation aside, we still must recognise the presence
of innumerable linguistic indicators of illocutionary force. 'Pragmatic'
guesses playa part, too, but if we had to rely on guesses, human commu-
nication would be much less successful, much less effective than it in
fact is. What is unsuccessful and ineffective are the models in terms
of which linguists have tried to analyse illocutionary forces. But the
solution is to abandon those inadequate models, and to replace them
with better ones, rather than to wash one's hands of any illocutionary
analysis, and to throw it out of linguistics altogether.
This is, then, the main thesis of this chapter: the supposed indetermi-
nacy of illocutionary forces is largely an artefact of inadequate syntactic
and semantic analyses.
1.1. Illocutionary forces as bundles of components
Consider again a sentence such as: Come here! If we analyse it as 'I
order you to come here' then we are indeed over-specifying its illocu-
tionary force. The speaker may have been 'asking' rather than 'order-
ing', or may have been simply 'telling' the addressee what to do. By
spelling out the illocutionary force as 'I order you' rather than as 'I ask
you' or 'I tell you', the linguist would be acting in an arbitrary and, one
might say, irresponsible fashion. Thus, when Leech (1983: 175) claims
that illocutionary force "is more subtle than can be easily accommodated
by our everyday vocabulary of speech act verbs" and that it cannot "be
200 The semantics of illocutionary forces
adequately captured by reference to such categories as offers, sugges-
tions, and statements" (1983:156), one can only agree with him.
In my view, however, it doesn't follow from this that illocutionary
force is necessarily 'indeterminate' or that "it must be studied in part
in non-categorical, scalar terms" (Leech 1983: 175).
We can't tell if an imperative utterance such as Come here! stands
for an order, or a request, or a command, but we can tell that it conveys
the idea which can be spelt out as 'I say: I want you to come here'. The
point is that orders, commands, and requests have something in common
(and of course this is why they can all be enacted by means of the same
grammatical category: the imperative). We don't have to say that Come
here! is an order (which would be arbitrary), or that it is ambiguous
between an order, a command and a request. Instead, we can extract the
semantic common denominator of these different interpretations. This
can be done precisely by means of the formula 'I say: I want you to do
it'. Contextual or suprasegmental clues may provide additional informa-
tion, but the core is signalled by the construction itself. The illocutionary
force of the utterance in question is quite 'determinate', but it can only
be captured in a framework which operates with sufficiently fine-grained
components.
What applies to the alleged 'indeterminacy' of illocutionary forces
applies also to their alleged 'scalar variability': both are artefacts of an
inadequate analytical framework. Consider, for example, the following
statement (Leech 1983: 175): "The difference between 'ordering' and
'requesting' is partly a matter of the scale of optionality (how much
choice is given to h), and the difference between 'requesting' and 'offer-
ing' is a matter of the cost-benefit scale (how far is A to the cost/
benefit of s/h)." (h stands for hearer, s for speaker, and A for act.)
But clearly, the difference between 'order' and 'request' or between
'request' and 'offer' can be represented by means of discrete illocution-
ary components. There is no need to invoke any 'scalar variability'. An
order includes a component which can be stated, roughly, as follows: 'I
think you have to do what I say I want you to do'; a request includes the
component 'I think: you don't have to do what I say I want you to do'; an
offer includes components such as 'I think you might want this', 'I think
this would be good for you'. (For a detailed discussion of these and
over two hundred other speech acts, see Wierzbicka 1987.)
Discussing 'whimperatives' such as Could you be quiet?, Bach and
Harnish (1982) consider two possible analyses: (1) The sentence is
ambiguous between a question and a request (Sadock 1970); (2) The
Are illocutionary forces indeterminate? 201
sentence is both a question and a request (Searle 1975). They conclude
that "the data suggest that whimperatives pattern in part like questions,
in requiring verbal responses, and in part like imperatives, in requiring
action for compliance. The conventionality thesis [i.e. Searle's analysis],
which has them being both, is better able to accommodate such facts
than the ambiguity thesis [i.e. Sadock's thesis] is." (Bach - Harnish
1982:186).
In other words, the choice is this: should mules be regarded as being
sometimes horses and sometimes donkeys or as being always both
horses and donkeys?
Not surprisingly, both analyses run into insuperable difficulties. As a
result, Bach and Hamish feel forced to abandon them both and to pro-
pose a third one. What is it? That mules are always ... horses! (i.e. that
whimperatives are always questions!) But since the obstinate mules don't
want to behave like horses and, for example, since they take a pre-verbal
(clause-internal) please, which questions generally don't do:
Could you please be quiet?
*How old please are you?
they therefore get condemned as 'bad horses', malformed horses, misbe-
having horses; and whimperatives, such as Could you please be quiet!
get condemned as ungrammatical!
What I propose is this: mules are not sometimes horses and sometimes
donkeys; and they are not both horses and donkeys; they are neither
horses nor donkeys; they are mules. Being mules, they are similar to
both horses and donkeys. In describing mules we must show in what
respects they are similar to horses and in what respects they are similar
to donkeys.
In what follows, I will discuss a number of syntactic constructions
which encapsulate specific illocutionary forces, trying to demonstrate
that although these forces can't be stated by means of single speech act
verbs, they can be stated with full precision by means of bundles of
components (cf. Wierzbicka 1972, 1977, 1980; cf. also Chapter 5 above).
I would add that by proposing a mode of analysis which requires
decomposition of illocutionary forces into illocutionary components I
am not proposing that the apparatus of linguistic description should be
further complicated or added to. On the contrary. Decomposition of
speech acts into illocutionary components is a necessary part of linguis-
tics anyway - if lexicography has anything to do with linguistics at all.
To state the meaning of verbs such as order, request, warn, threaten or
202 The semantics of illocutionary forces
offer we have to isolate their semantic components anyway. I suggest
the following mode of analysis (for justification and discussion see
Wierzbicka 1972, 1977, 1980, 1987; see also Chapter 5, section 1):
I order you to do this (X)
I say: I want you to do this (X)
I say this because I want you to do it
I think:
you have to do it because of this
you will do it because of this
I ask you to do this (X)
I say:
I want you to do this (X)
it will be good for me
I say this because I want you to do it
I think:
you don't have to do it
I don't know if you will do it
I suggest that you do this (X)
I say: I think it may be good if you do this (X)
I say this because I want you to think about it
I think: I don't know if you will want to do it
To decompose the verb order into its semantic components is the
same thing as to analyse the illocutionary force of the speech act 'order'
(pace Searle 1979). There would seem to be little point, therefore, in
assuring ourselves that we as linguists don't have to engage in an
analysis of illocutionary forces because somebody else (a conversational
analyst) will do it for us. We have to do it, as long as lexical semantics
is a part of linguistics. And if decomposition of illocutionary forces
into components solves at the same time the insuperable syntactic and
semantic difficulties that the 'performative hypothesis' ran into, this is
a sheer bonus.
1.2. Illustration: the discrete and determinate character of
'whimperatives'
Consider, for example, the following characteristic remark (Comrie
1984b:281): " ... 'Would you open the door' is normally going to be used
Are illocutionary forces indeterminate? 203
as a request, perhaps even as a command. The usual response would be
to carry out that request. But notice that a possible response is 'No'."
Comrie's hesitation between 'request' and 'command' makes it sound
as if the pattern in question were somehow 'indeterminate'. In fact,
however, the apparent indeterminacy stems entirely from the inappro-
priateness of the descriptive categories used, and - as Comrie's
subsequent discussion suggests - the illocutionary force of the Would
you pattern is quite specific and includes a mixture of 'directive' and
'interrogative' components. This can be illustrated with the following
exchange:
Standing on the porch, before she rang the bell, Miss Ellis took
out a comb. 'Would you try to pull this through your hair?'
Gilly shook her head. 'Can't.'
'Oh, come on, Gilly - '
'No. Can't comb my hair. I'm going for the Guinness Recordfor
uncombed hair.'
'Gilly, for pete's sake ... ' (GH)6
Evidently, the speaker does want something, and her subsequent
protestations (Oh, come on, Gilly - , for pete's sake ... ) make this
abundantly clear. At the same time, the directive is phrased tentatively,
it shows respect for the addressee's personal autonomy, it solicits a
verbal response (as well as an action), and implicitly acknowledges the
possibility of obstacles, counter-indications, and the like ('would you
do it if - if I said I wanted it?').
To show the full force of an illocutionary pattern such as Would you,
we need to reveal all the illocutionary components encapsulated in it.
This calls for a painstaking qualitative analysis, taking into account
both the form and the communicative range of the pattern in question.
It is an illusion to think that instead of engaging in a qualitative analysis
of this kind we can simply draw some 'continuum' of illocutionary
forces, such as that advocated by Giv6n (1984:249 and 1989:153) as
the "continuum ... between prototypical imperative and interrogative
speech acts":
Most prototypical imperative
a. Pass the salt!
b. Please pass the salt.
c. Pass the salt, would you please?
d. Would you please pass the salt?
204 The semantics of illocutionary forces
e. Could you please pass the salt?
f. Can you pass the salt?
g. Do you see the salt?
h. Is there any salt around?
i. Was there any salt there?
Most prototypical interrogative
According to Givan (1989: 154), three socio-psychological dimensions
underlie this continuum: "(a) the power (authority) gradient between the
speaker and the addressee, (b) the degree of the speaker's ignorance
concerning a state of affairs about which he wishes to learn, and (c) the
degree of the speaker's sense of urgency or determination vis-a-vis the
attempted manipulation. All these parameters are scalar, thus represent-
ing a multidimensional space."
Quite apart from the fact that these three hypothetical dimensions are
undefined, ad hoc, and arbitrary (for example, why 'urgency'? does the
bare imperative always imply 'urgency'?) their 'scalar' character is
clearly an artefact of Givan's presentation. For example, the difference
between Pass the salt and Please pass the salt is perfectly discrete,
and if Givan's methodological framework doesn't provide him with
adequate tools for spelling it out, it is a feature of the framework, not
of the data under discussion.
Being an artefact of Givan's analysis, rather than a reflection of the
nature of linguistic communication, the 'scalar description' of illocution-
ary forces could nonetheless, in principle, serve some useful purpose in
applied linguistics. I doubt, however, that it does that either. Consider,
for example, the situation of a Russian learner of English who is told that
the Would you pattern is number 4 from the top on the continuum from
the 'most imperative' to the 'most interrogative' sentences. What use
would this be to him or her? As Comrie points out, Russian doesn't
use the Will you or Can you patterns in the way English does, and the
same applies to the Would you pattern:
. .. in English one polite way of getting someone to do something is by
asking a yes/no question using either some form of 'will' or some form of
'can'. In other languages, that's not conventionalised. If you tried it in
Russian, the reaction would be 'What's this guy trying to do?' (Comrie
1984b:282)
Givan did not put the Will you and Can you patterns on this continuum,
so one doesn't know what slots he would have assigned to them: 3 and
4? 4 and 5? perhaps 3 and 5? Whichever numbers he might decide to
Are illocutionary forces indeterminate? 205
assign to these patterns, their value to a language learner would be
limited, and they could not substitute for a qualitative description of
the meanings and uses of the construction types.
Furthermore, it would not always be easy (or possible) to rank differ-
ent types of utterances 'from the most imperative to the most interroga-
tive'. For example, Giv6n places Would you above Could you, and
presumably he would opt for placing Will you over Would you. But
how would one rank, for example, Will you please with respect to Would
you (without please), or Would you please with respect to Could you
(without please)?
Incidentally, Comrie's comment linking some forms of will and can
whimperatives with politeness is also misleading, given the fully felici-
tous nature of English sentences such as Will you please shut up! or Will
you bloody well hurry up! (Williamson 1974:56; cf. Chapter 2 above).
But if we reveal the different configurations of illocutionary patterns,
this may be practically helpful, as well as theoretically illuminating. In
particular, it will show which patterns expect a verbal response and
which don't; which patterns assume (according to the speaker) that the
addressee 'has to do' what the speaker wants and which don't; which
show that the speaker feels bold enough to say 'I want you to do it'
and which reveal the speaker's reluctance to go on record as having
said this. And so on.
Consider, for example, the relationship between the three English
constructions: the bare imperative, the Will you pattern, and the Would
you pattern. The semantic structure of the bare imperative (of action
verbs) can be represented as follows:
Do this (X)
(a) I say: I want you to do this (X)
(b) I say this because I want you to do it
(c) I think: you will do it because of this
This is closely related to the explication proposed earlier for orders, but
there is one important difference: while both orders and bare imperatives
confidently expect that the speaker's 'want' will be complied with,
orders include one additional component: an assumption that the
addressee has to do what the speaker wants. For example, if one shouts
to someone unaware of a sudden danger: Step back!, this is not an order,
and there is no assumption that they have to comply with the speaker's
will, but there is an expectation that they will obey.
206 The semantics of illocutionary forces
The semantic structure of the two 'whimperatives' can be portrayed
as follows:
Will you do X?
(e.g. Will you please shut up? Will you pass the salt?)
(a) 1 say: 1 want you to do X
(b) 1 say this because 1 want you to do it
(c) 1 don't know if you will do it
(d) 1 want you to say if you will do it
Would you do X?
(e.g. Would you try to put this through your hair?)
(a) 1 say: 1 would want you to do X
(b) 1 say this because 1 want you to do it
(c) 1 don't know if you would do it if 1 said 1 wanted you to do it
(d) 1 want you to say if you will do it
Component (b), which spells out the illocutionary purpose, is the
same in both cases ('I say this because 1 want you to do it'), and it links
the two patterns in question with bare imperatives (Do X). Component
(d), which solicits a verbal response (in addition to a non-verbal one) is
also the same in both cases. But the most overt component, which can
be called the dictum (a), is different in each case: in the Will you pattern
it is the same as in the case of the straight imperative ('I say: 1 want
you to do it'), whereas in the Would you pattern it appears to be more
tentative ('I say: 1 would want you to do it'). Furthermore, while neither
of the two patterns is totally confident about the outcome, the Will
you pattern constitutes a more forceful attempt to ensure the desired
outcome. This difference is reflected not only in the (a) component
(the dictum) but also in the ignorative component (c): compare 'I don't
know if you will do it' with 'I don't know if you would do it if 1 said
1 wanted you to do it'.
This analysis accounts, 1 think, for both the similarities and differ-
ences in the way the two patterns are used. For example, according to
my informants the sentence Would you bloody well hurry up? is less
felicitous than Will you bloody well hurry up? - presumably, because
the forceful, uninhibited nature of the verb shut up can be seen as
clashing with the conditional and rather tentative character of the Would
you pattern. 1 might add that when my research assistant leaves time-
sheets for me to sign she usually puts on them a pencilled line "A(nna),
would you sign these?", and never "A, will you sign these?", which
Are illocutionary forces indeterminate? 207
would sound (in my judgement and in that of my informants) more
peremptory and less courteous.
Furthermore, the analysis proposed here accounts also for the formal
features of the patterns in question, and in particular, for their links
with both imperative and interrogative constructions; and with tagged
imperatives such as Do X, will you? and Do X, would you?, which will
be discussed later.
Another closely related pattern is Can't you ... ?, which suggests
even more irritation and dissatisfaction, as the example below illustrates
(note also the use of just and stupid):
'Can I come in?'
'No!' shrieked Gilly, then snatched open the door. 'Can't you
leave me alone for one stupid minute?' (OR)
Can't you just leave me alone? (OH)
I propose for this pattern the following explication:
Can't you do X?
I see you are not doing something that you should be doing
I feel something bad because of this
I think you should know: this is bad
I say: I want you to say if you can't do this thing
I think you can do it and don't want to do it
I say this because I want to say what I think and what I feel
The form of the patterns described above is not arbitrary, and it can be
partly understood in terms of 'conversational implicatures' and 'natural
logic' because it reflects some of the components of the semantic
(illocutionary) structure. It is, however, language-specific, and it is
partly 'arbitrary' (that is, unpredictable from a cross-linguistic perspec-
tive), because it doesn't reflect overtly all the components of the illocu-
tionary structure.
1.3. Syntax and illocutionary force
Consider utterances such as the following ones (Schreiber 1972:326;
Levinson 1983:256; Davison 1975:163).
Frankly, I lied.
John's at Sue's house, because his car's outside.
208 The semantics of illocutionary forces
Proponents of the performative hypothesis have often tried to account
for the syntax of such sentences by claiming that the adverb frankly
and the because clause modify the underlying performative verb tell.
However, as Levinson pointed out:
. .. there are significant semantic difficulties here. ... it is simply not
clear that the meanings of the relevant adverbs are indeed parallel in the
explicit performative, the (allegedly) implicit performative and the reported
performative usages:
(50) I tell you frankly you're a swine.
(51) Frankly, you're a swine.
(52) John told Bill frankly that he was a swine.
(Levinson 1983:255)
In prefacing one's sentence with frankly, as in (51), the speaker is not
commenting on her own comment while delivering it, as (50) would
suggest. Rather, she is saying two things, first preparing the addressee
for what is to follow:
I now want to say something frankly
I say: you're a swine
As for sentences with a because clause:
it is clear that the because-clause here does not in fact modify any implicit
I state or I claim, but rather an understood I know, as made explicit in
(59):
(58) I state John's at Sue's house because his car's outside.
(59) I know John's at Sue's house because his car's outside.
(Levinson 1983:257)
(Noticing John's car outside the speaker infers, in her own mind, that
John is at Sue's house. Her stating that John is at Sue's house follows
this inference, and may be due to a number of causes. The direct causal
links hold between noticing that A and saying that B.)
Perhaps the clearest evidence of the inability of the performative
hypothesis to deal with the kind of data it was originally meant to
account for comes from metalinguistic comments such as To sum up, To
change the subject or To cut a long story short. As Mittwoch (1977: 183)
pointed out, "it would be counter-intuitive to derive (29b) [To change
the subject] from: I tell you this in order to change the subject." One
says To change the subject to alert the addressee that a new topic is
being introduced; but what one says about this new topic may be said
Are illocutionary forces indeterminate? 209
in order to inform the addressee, to express a sudden thought, and so
on - not necessarily in order to change the subject.
I quite agree with Levinson, Mittwoch and others, that the attempts
to explain the interaction between syntax and illocutionary force on the
basis of the performative hypothesis were unsuccessful. But the alterna-
tive accounts, proposed by the opponents of this hypothesis such as
Leech (1983) or Bach - Hamish (1982), are hardly more convincing.
What this alternative account suggests is, to my mind, quite incredible:
that all sentences which signal overtly some aspect of their illocutionary
force, and which can't, therefore, be explained in terms of 'autonomous'
(non-illocutionary) syntax, are simply ungrammatical! They are of
course perfectly acceptable, and they are used all the time, but since
our favourite grammatical theories can't generate them, then they are
ungrammatical. Because "usability is not grammaticality, and acquiring
a use does not turn the ill-formed into the well-formed" (Bach -
Harnish 1982:225). In other words, if there is a conflict between
grammatical theory and plain facts - then too bad for the facts!
Among facts thus condemned by the grammatical theory are all the
'whimperatives', all the 'queclaratives', all sentences with so-called
'style disjuncts': in fact, it seems, the bulk of conversational English.
Thus it is 'ungrammatical' to say any of the following utterances (Leech
1983:193-194; Bach - Hamish 1982:219,230):
Can you please close the window?
Frankly, you bore me.
Who gives a damn about that?
Why don't you be quiet?
No smoking.
Two coffees, please.
Shut the window, can't you?
As Leech (1983: 195) puts it, "such sentences do not lend themselves to
generalisation in any kind of grammatical framework, being essentially
exceptions to general rules".
For my part, I would like to ask: what is so sacrosanct about those
'general rules'? If they don't fit the facts shouldn't we re-examine the
'rules', rather than condemn the facts?
In fact, it is not true that facts of this kind can't be accommodated
in any kind of grammatical framework. They can be accommodated in
a framework which derives surface structures from semantic structures
couched in terms of illocutionary components. For example, while it is
210 The semantics of illocutionary forces
true that sentences with the illocutionary adjunct frankly can't be
plausibly derived from I tell you frankly, they can be plausibly derived
from the following structure:
1 now want to say something frankly
1 say: X
An infinitive clause such as to change the subject can't be plausibly
derived from I tell you this in order to change the subject, but it can
be plausibly derived from a structure where it will modify one of the
illocutionary components: 'I say this because ... " 'I know this because
... " 'I think this because ... " and so on.
Generally speaking, the main mechanism which operates in deriving
surface structures from semantic structures is that of abbreviation, or
'deletion'. A 'whimperative' sentence, which combines in its surface
structure imperative and interrogative features, must be derived from an
underlying structure which contains, among others, components such as
'I want you to do something', and 'I want you to say something'. But the
full meaning of a particular illocutionary form (for example the full
meaning of Why don't you do X? or of How about X?, or of Why do X?)
is very complex, and it is language-specific. The view that a sentence in
a frame such as Why don't you is ungrammatical, and that its force can
be guessed on the basis of "the general principles of rational, purposive
human behaviour" (Leech 1983:195), is singularly unhelpful, and indeed
perverse - especially from the point of view of a second language
learner in whose native language a literal equivalent of this frame can't
be used with anything like the force that it has in English. An immigrant
learning English might guess that an utterance in the frame Why don't
you will have an interrogative component in its meaning, but can't
possibly guess its full meaning. This full meaning has to be stated for
the learner. It is surely the responsibility of the linguist who wants to
describe English to say what this meaning is.
More whimperative constructions 211
2. More whimperative constructions
2.1. Why don't you do X (tomorrow)?
A sentence such as Why don't you play tennis any more? can be a
straightforward question. If, however, a sentence in the frame Why
don't you refers to a specific (non-habitual) action and has a future time
reference, as in:
Why don't you go and see a doctor tomorrow?
then the sentence cannot be simply a question: it must convey the
assumption that it would be a good thing for the addressee to do the
thing mentioned. Green (1975:127) has pointed out that the sentence:
Why don't you be quiet! is an unambiguous 'whimperative', whereas the
sentence: Why aren't you quiet? is an unambiguous question. I think
that this observation can be generalised: not just be, which differs from
the present tense form are, but any infinitive combined with why don't
you will be interpreted as a 'whimperative', provided that the sentence
has a specific time reference.
In a particular context, it is often fairly clear that a sentence in the
Why don't you frame is in fact meant as an invitation, or as an offer, or
as a suggestion, or a request, for example:
'Did you do division with fractions at Hollywood Gardens?'
Gilly shook her head. ...
'Why don't you bring your chair up to my desk and we'll work
on it?' (GH)
[offer of help]
Why don't you bring me a wild one, Miss Gilly? I need to wake
up that fifty-year-old senior citizen I've got for a son. (GH)
[request for a necktie]
Why don't you just go up and finish your nap, Mr. Randolph? I
feel bad waking you up like this. (GH)
[solicitous suggestion]
Tell you what, why don't you stay the weekend? (GP)
[suggestion/invitation]
212 The semantics of illocutionary forces
But even in such relatively clear cases, a measure of 'indeterminacy'
remains. For example, the invitation and the offer cited above could
also be reported with the verb suggest rather than with invite and
offer. It is important to realise, however, that this 'indeterminacy' is
a function of the descriptive framework which forces the analyst to
choose between invite, offer, suggest and request. Utterances in the
frame Why don't you do X? can be reported in a number of different
ways because English has a number of illocutionary verbs which are
compatible with the illocutionary force encoded in this construction.
In choosing a particular verb (request, suggest, or whatever) the re-
porter imposes a certain interpretation on the original utterance, and can
choose one of a number of interpretations compatible with the force
signalled by Why don't you. In doing so, the reporter adds to what is
encoded in the construction itself. This is, then, where the 'indetermi-
nacy' lies: in the range of possible interpretations, which can be signal-
led by a range of reporting verbs. But the force of Why don't you is
quite determinate. It can be stated as follows:
Why don't you do X?
(a) I say: I want you to say, if you can, why you/we shouldn't
do X
(b) I think you can't (say why)
(c) I think it will be a good thing if you/we do it
(d) I say this because I would want you to do it
Since it is the task of linguistics to pair surface structures with the
meanings encoded in them, it is plainly the task of the linguist to spell
out, in particular, the meaning of a frame such as Why don't you - and,
I believe, this task can be carried out in the way indicated above.
The formula postulated here can be validated in terms of meaning,
in terms of form and in terms of use. The meaning can be verified
intuitively, since the metalanguage used in the formula is derived from
natural language. The form can be explained as derived by mere dele-
tions from the underlying structure. (Of course, to say that the rules
linking semantic representations to surface forms are for the most part
'mere deletions' doesn't mean that they don't have to be described, and
accounted for in terms of more or less general principles.) And various
aspects of use can be accounted for in terms of individual components
either present or absent in the postulated semantic structure. For ex-
ample, the fact that the frame in question can't take a pre-verbal please
can be explained by the absence from the underlying structure of the
More whimperative constructions 213
component 'I want you to do it', which would be posited for 'whimpera-
tives' such as Could you do X? but not for 'impositives' such as Why
don't you do X:
Could you please be quiet.
*Why don't you please be quiet.
It is particularly interesting to note that, although more tentative than
a straight imperative, the Why don't you pattern does not have to be
particularly 'polite'. For example, it is perfectly felicitous in curses,
such as Why don't you all go to hell! (Hibberd 1974:199). But a curse
of this kind - contrasted with the imperative Go to hell! - suggests
a somewhat impotent exasperation rather than self-confident anger.
(For further discussion of the formula proposed here see the following
subsections.)
2.2. Why do X?
Now consider sentences such as the celebrated:
Why paint your house purple?
(Gordon - Lakoff 1975). The long controversy over whether a sentence
of this kind is a question or a criticism seems to me rather futile. In
fact, it partakes of the nature of both question and criticism, and its
illocutionary structure simply can't be captured accurately in terms of
global categories such as ask and criticise. What is needed are more
fine-grained components, such as:
Why do X?
(a) I say: I want you to say, if you can, why you/we should do X
(b) I think you can't (say why)
(c) I think: X is not a good thing (to do)
(d) I say this because I want to say what I think
I maintain that the syntactic construction Why do X? always carries these
components with it, not only in those sentences where the content
supports this interpretation but also in those where it doesn't. For
example, the sentence:
Why paint your house white?
214 The semantics of illocutionary forces
implies exactly the same attitude as the sentence with the word purple
in lieu of white. This means that the construction itself provides unambi-
guous grammatical clues to the illocutionary force. But this force cannot
be stated as simply 'I ask' or 'I criticise', or even as 'I ask and I criticise'
(a la Sadock 1974). It can only be stated in terms of a set of illocution-
ary components.
It is worth adding that the construction in question has a more or less
symmetrical counterpart of the form Why not do X?, as in the following
example:
Why not have him come here? (Why go all the way to his place?)
The illocutionary force of this 'twin' construction differs from that
of Why do X? in terms of evaluation and also in terms of illocutionary
purpose. Why do X? implies that X is not a good thing to do; Why not do
X?, on the other hand, implies that X is a good thing to do. The former
is a kind of tentative criticism; the latter, a kind of tentative suggestion.
The former can refer to faits accomplis, but the latter has a future orien-
tation. For example, if a house has already been painted, and cannot be
repainted, one can still say
Why paint the house purple?
But if one says:
Why not paint it purple?
one is implying that it is not too late to carry out the 'good idea' in
question. Searle (1975:69) insists that a sentence in the frame Why not
do X? can be a straightforward question 'without necessarily being
also a suggestion'. As I see it, it is true that sentences of this kind don't
have to be suggestions, but it is not true that they can be straightforward
questions, because they will always contain some component such as 'I
think that it would be a good thing to do X'.
I suggest that the illocutionary force of the Why not frame can be
represented as follows:
Why not do X?
(a) I say: I want you to say, if you can, why you/we shouldn't
do X
(b) I think you can't (say why)
(c) I think it would be a good thing if you/we did it
(d) I say this because I want you to think about it and to say
what you think about it
More whimperative constructions 215
Comparing the explications for the frames Why don't you do X? and
Why not do X? one will see that they share a number of components, and
that they are both compatible with the meaning encoded in the verb
suggest. One will also see that they differ in ways which make the
former, but not the latter frame, compatible with the meaning of request.
Thus, the first three components of this formula (a, b and c) are the same
as those posited for the frame Why don't you? But the component (d)
postulated for Why don't you? ('I would want you to do it') has been
omitted from the explications of Why not do X?, to show that this time
the speaker presents himself as totally disinterested. The illocutionary
purpose of each of the two constructions is also different: in the case of
Why don't you? sentences the speaker is attempting to cause the action
to happen (if the addressee is in favour); but in the case of Why not
do X? sentences the speaker is merely trying to cause the addressee to
make up his/her mind.
2.3. How about X?
How about sentences are usually used for making suggestions, offers or
invitations (but not, for example, orders or instructions):
How about a drink?
How about a movie?
How about you and me doing a little red-hot reading after
supper? (GH)
How about dinner at my place?
How about going to Sydney?
One can't claim, however, that sentences of this kind can be represented
as derived from or equivalent to I suggest (that we see) a movie, I offer
you a drink or I invite you to dinner, because it would involve a
great deal of arbitrariness to decide which performative verb to use in
each case. And to suggest that all How about sentences are multiply
ambiguous (not simply ambiguous but multiply ambiguous) is almost
tantamount to admitting their indeterminacy.
But if one admits this then one is left with an analysis devoid of
explanatory power; because if How about sentences are either multiply
ambiguous or indeterminate then why is their range of possible illocu-
tionary forces so strangely limited? Why can't they be used, for example,
for ordering, commanding, begging, imploring, or giving permission?
216 The semantics of illocutionary forces
The 'mystery' is easily solved, however. If How about sentences can be
used for suggesting, offering or inviting, but not for ordering or begging,
it is because the former three acts are compatible with the components
encoded in English by the How about construction. These components
can be spelt out as follows:
How about doing X?
(a) I think you might want to do X
(b) I say: I want to know 'how you would feel about it'
(c) I would want it to happen if you wanted it
(d) I don't know if you would want to do it
(e) I say this because I want you to think about it and to say
what you think about it
The action is presented as potentially desirable, but the speaker is not
as confident on this point as in the case of Why don't you sentences.
Furthermore, the speaker doesn't advocate the action as desirable from
his/her own point of view, but rather invites the addressee to form and
express a view. The speaker is not saying, therefore, 'I think it would be
a good thing if it happened', but rather, 'I think you might want to do it'.
There is no assumption, therefore, that the action should be carried out.
Nonetheless, the speaker does indicate a personal interest in the action,
but as in the case of Why don't you sentences, this personal interest
is expressed in a conditional form ('I would want it to happen if you
wanted it').
A How about sentence is more 'open-minded' or open-ended, there-
fore, than a Why don't you one. The speaker wants to hear if the ad-
dressee wants to carry out what has been proposed as potentially desir-
able, but is not pressing the addressee in any way. (It is interesting to
note in this connection that while the sentence Why don't you all go
to hell! is felicitous, the sentence How about you all going to hell!
would be comical.)
Since the How about pattern is used commonly for offers and sugges-
tions it might seem that it presents the action in question as 'good for the
addressee', and that its explication should therefore include a component
of the form 'I think it would/might be good for you'. But in fact, this
pattern can also be used for requests (though not for orders), as in the
following examples:
'How about giving me a hand with this salad?'
'No.'
'Oh.' (GH)
More whimperative constructions 217
Supper's' bout ready. How about going next door and getting
Mr. Randolph? He eats here nights.
The word No was just about to pop out of Gilly's mouth, but
one look at Trotter's eyes, and she decided to save her fights
for something more important. (GH)
The important feature of such requests is that they are counting on the
addressee's willing cooperation. This is compatible with the component
'I think you might want to do X', but not with 'I think it would/might
be good for you.'
In an insightful discussion of How about sentences in English, Shopen
(1974:794) argues that the pattern in question "should be viewed as
having status in the grammar in its own right". I would entirely agree
with this claim, in view of the unique illocutionary force signalled by
this pattern. I could not fully agree, however, with the concomitant claim
that "there is no non-elliptical source ... from which this pattern could
be derived that has the same semantic properties".
An explication of the kind proposed above can be regarded as just
such a 'non-elliptical source' - at least in the sense that it constitutes
a non-elliptical paraphrase (or 'paralocution', cf. Boguslawski 1981a)
which spells out the unique illocutionary force of the How about pattern.
To say this is not to deny that the expression How about can also be
used in 'pure questions'. But as Shopen rightly recognised, apart from its
use in pure questions it can also be used with an illocutionary force
different from that of questions. I suggest that this illocutionary force
can be fully spelt out in a non-elliptical formula which takes the form
of a unique series of illocutionary components (unique, and yet
sufficiently overlapping the formula for 'pure questions' as to account
for the formal link between them).
It should be added that the 'non-elliptical' formula proposed here
includes elements other than those postulated as universal semantic
primitives (for example 'how', 'feel'); and also, that phrases such as
'feel about' do not belong to the proposed semantic metalanguage. In
this sense, the explication sketched here cannot be regarded as the true
'non-elliptical source' of the expression in question. I have kept, how-
ever, the phrasing 'how you would feel about it' to show how the 'how
about' pattern can be seen as related to (and as an abbreviated form of) a
'deeper' (intermediate) level of semantic representation (cf. Wierzbicka,
in press b). The paralocution 'how you feel about it' is still semi-idio-
218 The semantics of illocutionary forces
matic, and at a deeper level of analysis it should (and could) be replaced
with a non-idiomatic configuration of universal semantic elements.
3. Additional remarks on the explication of
illocutionary forces
Several formulae explicating illocutionary forces have been introduced
so far, and many more are going to follow. It may be useful to add a
few comments on the methodology of semantic explications at this stage,
for the reader may still feel puzzled: how does one know what com-
ponents should be posited for a given construction? and how does one
prove that the components proposed have not been arbitrarily chosen?
I reply that one proceeds, essentially, by trial and error. The goal is
this: to propose for each construction a minimal set of components
which will jointly account for all the aspects of its use. As a research
strategy, a contrastive approach is most fruitful: if one tries to model
several closely related constructions at the same time, trying to capture
both the similarities and the differences between them, one has the best
chance of providing the most accurate 'portrait' of each illocutionary
force. However, the ultimate goal consists in explicating, as fully and
accurately as possible, each illocutionary force in its own right - not
in providing some sort of differential schema for the whole lot. Above
all, one must resist the temptation of imposing some 'system' on the
empirical reality of illocutionary forces, which can be as capricious,
idiosyncratic and asymmetric as any other semantic structures. To be
sure, a great deal of symmetry and order is also likely to be discovered,
but the exact proportion of symmetry and asymmetry, of 'system' and
idiosyncracy has to be revealed through empirical research, not decided
about on an a priori basis.
For example, if we decide to posit the component 'I think you might
want to do it' for the frame How about (How about a beer, How about a
movie, and so on) but not to posit it for the frame Why don't you (Why
don't you shut up), it would be a grave error to posit for the latter
frame a negative component 'I don't think you might want to do it', or
'I don't say that 1 think that you might want to do it'. There is an
essential difference between the absence of a component and the pres-
ence of its opposite.
Selected conversational strategies 219
Semantic explications are difficult to justify fully (within a single
study, or even a single volume), because ultimately, the only way to
justify them is by refuting alternative formulations. Each explication
proposed in the present chapter has gone through a number of
versions, and in each case dozens of competitors have been considered
and discarded. This is of course no guarantee of perfection. But it
would be an illusion to imagine that there is (or could be) some simple,
mechanical procedure for evaluating semantic explications. The only
way to challenge them is to engage in the laborious process of devising
competing analyses, and of defending them, point by point, against
attempts at refutation.
The present study of various sentence types and grammatical forms,
and the illocutionary forces they encode, may be seen as a first step in
this process. I try to propose for each construction a set of (illocutionary)
semantic components which can account for its use. Once proposed,
they are accessible to discussion, revision, refutation or verification.
4. Selected conversational strategies
This section examines several examples of conversational frames used
for different illocutionary purposes: making suggestions, introducing
information, and indicating the speaker's attitude or opinion. An ex-
plication of the illocutionary force of each construction is proposed.
4.1. Tell you what, Sf
Acts similar to suggestions can also be signalled in English by means
of the frame (I) tell you what, as in the following example, (quoted
earlier for a different purpose):
Tell you what, why don't you stay the weekend (GP)
An utterance in this frame could be reported with the verb suggest ('he
suggested that she stay the weekend'), just as Why don't you utterances
and How about utterances can. In fact, in the example given above Tell
you what and Why don't you are used jointly.
220 The semantics of illocutionary forces
But Tell you what implies that the idea has just occurred to the
speaker, whereas How about or Why don't you could voice an idea
which the speaker has been nurturing for a long time. Furthermore, Tell
you what implies that the sudden idea which has just occurred to the
speaker will constitute a solution to a problem of current relevance. The
word what seems to echo a question which has been occupying the
interlocutors' minds (what should we do?). It offers an answer to that
implicit question: (I) (will) tell you what I think (we should do). No
such implications are included in the formally similar expression I
(will) tell you something.
The reference to the first person should of course be interpreted
broadly. As pointed out to me by Tim Shopen, one could say, for
example:
Tell you what, she could try auditing a course.
But even here, the implication seems to be that the speaker and the
addressee are somehow responsible for the person in question.
I think the following formula can be proposed for the frame Tell you
what (for a different analysis, see Fillmore 1984):
Tell you what, Sf
I think I now know what you/we should do
because I now think something (I didn't think it before)
I want you to know it
I say: I think it will be good if you/we do something (S)
I say this because I want you/us to do it if you think the same
4.2. Do you know, S?
Do you know, he has started a new poem? (NI)
Do you know sentences are related to Tell you what sentences, in so far
as both types signal 'news', and moreover, news that the addressee can
be expected to be interested in. In the case of Tell you what sentences,
the 'news' is directly relevant to the addressee, because the speaker has a
'good idea' concerning the addressee or someone close to the addressee
('it would be good if you/we did X'). No such direct relevance is
implied by Do you know. What is implied by the latter frame is
simply interest: 'I think you would want to know this'. The full semantic
formula would read:
Selected conversational strategies 221
Do you know, S?
I say: I want to say something to you (S)
I think you would want to know it
I think you wouldn't have thought this
I say this because I want you to know it
I think you will say something because of this
I want you to say something because of this
An interesting formal feature of Do you know sentences used in this
sense is that they don't allow an explicit complementiser: Do you know
THAT S indicates that the addressee may already know the news.
Do you know, I had a dream about you last night.
?Do you know that I had a dream about you last night.
Sentences in the form Do you know that may also introduce interesting
and unexpected news, but they have to have an 'ignorative' component
'I don't know... ' (cf. Wierzbicka 1980:315):
Do you know that X?! =>
'I don't know if you know it'
The contrast in illocutionary force between Do you know (*that) S?
and Do you know that S? can be illustrated with the following pair
of examples:
Do you know, for years I used to dream that he'd caught us in
bed together? You and me. Even after we were married. (GP)
'It's absolutely disgusting!'
'Do you know, Denise', Gavin said, 'that the foolish folk
of Staunton might well, and in a deeply depressing majority,
find it more disgusting that your delightful nipples are visible
through your charming jersey.' (GP)
In the first example, the speaker knows that the addressee cannot possi-
bly know the content of the complement clause, and he introduces this
clause by Do you know, without that; but in the second, the speaker
cannot have the same certitude, and uses the variant with that (Do you
know that).
What holds for the frame Do you know that holds also, mutatis
mutandis, for the frame Did you know that. A sentence such as:
222 The semantics of illocutionary forces
Did you know that the Spanish anarchists once passed a resolu-
tion saying that any woman who excited a man's desire had a
moral obligation to satisfy it? (GP)
could be meant as a piece of news, rather than as a genuine question, but
it would still have an 'ignorative' component, 'I don't know if you know
it'. But the same sentence introduced by the frame Do you know (*that)
would imply that the speaker assumes the addressee doesn't know. And
a piece of news which couldn't possibly be known to the addressee
would have to be introduced by Do you know (*that), not by Did you
know that:
*Did you know that I had a dream about you last night?
4.3. Don't tell me Sf
Eleanor: (angrily). Michael! Don't tell me you're becoming
jealous of John again! (WE)
Unexpected news can also be indicated by the frame Don't tell me. This
time, however, what is said is new to the speaker, not to the addressee.
This means that the speaker discovers something new and unexpected,
and voices this discovery, with disbelief and disapproval. Usually, the
discovery concerns the addressee,. and is based on something that the
addressee has said, (as in the example above), but it doesn't have to be
so. For example, catching a glimpse of a female acquaintance one might
exclaim to somebody else:
Don't tell me she's pregnant again!
Similarly, having planned an excursion and having made no provision
for bad weather one could exclaim:
(Oh no!) Don't tell me it's raining!
But one would have to be a masochist, as well as a pessimist, to be
able to exclaim:
Don't tell me the weather's fine!
As a first approximation, 1 propose to spell out this illocutionary force
by means of the follolwing formula:
Selected conversational strategies 223
Don't tell me S!
I see something now
I think I have to think S because of this
I don't want to think this
I wouldn't have thought I would see this
I feel something because of this
I think one cannot say that this is not a bad thing
I want you to say something because of this
I say: I don't want you to tell me S
I say this because I want to say what I think
The discovery, the disbelief, the negative evaluation and the conse-
quent feeling link the expression Don't tell me S with the exclamation
Oh my God (see below, section 20). But Don't tell me, unlike Oh my
God, leaves some room for doubt, and it includes an appeal to the
addressee: 'I want you to say something because of this', 'I don't want
you to tell me S'.
4.4. How many times have I told you (not) to do Xl
Even stronger disapproval than that embodied in Don't tell me is of
course signalled by expressions of direct rebuke (reproach, reprimand
etc.) such as How many times have I told you (not) to do X. In this
case, the disapproval has to be directed at the addressee (rather than
at a third person), and it is accompanied by an attempt to make the
addressee feel bad, because of what he or she has done. The full
illocutionary force can be spelt out as follows:
How many times have I told you not to do X!
I see that you did X
I think you know that you shouldn't do this
because I told you many times
I say: I want you to say how many times it was, if you can
I think you can't, because it was very many times
I feel something bad because of this
I say this because I want you to feel something bad
The expression 'I feel something bad', included in this explication and
in many others, sounds no doubt somewhat naive and unidiomatic, but
if we specified the 'bad feeling' in question as displeasure, annoyance,
224 The semantics of illocutionary forces
irritation, anger or anything else, we would be acting arbitrarily. On
the other hand, to say that the component in question is 'indeterminate'
(between annoyance, anger, etc.) would be just as unjustified as to
say that How many times have I told you as a whole is indeterminate
between rebuke, reproach, reprimand, scolding and whatever.
4.5. Who's talking about doing X?
Who is talking about getting married? You talk as if marriage
was the only alternative. Don't you ever listen? (GP)
This is another construction expressing the speaker's annoyance with the
addressee. In this case, the annoyance (or, more generally, 'bad feeling')
must be caused by something that the addressee has just said, and the
nature of the offense is quite specific (although it is not specified on the
surface of the sentence): the addressee has attributed to the speaker an
intention which in fact the speaker doesn't have.
The expression Who's talking about (or of) doing X? is used as a
kind of angry disclaimer, protesting the addressee's mistaken assump-
tion. The full illocutionary force can be spelt out as follows:
Who's talking about doing X?
I think you think that I want to do X
I think you think this because I said something now
I feel something bad because of this
I say: I don't know who is talking of doing X
I say this because I want you to say it (who) if you can
I think you can't
I don't want you to think that I want to do X
5. Tag questions
5.1. Tags with declarative sentences
To take another set of examples, I will tum now to a number of differ-
ent tag question constructions (cf. Chapter 2, section 3.3). First, let us
consider declarative sentences with an opposite polarity tag such as:
Tag questions 225
Maria is Italian, isn't she?
At first sight, it might seem that sentences of this kind express the
speaker's uncertainty, and seek confirmation from someone who might
know what the truth is. But clearly, it is not knowledge or truth that
the speaker is seeking in examples such as the following ones:
Maria is very nice, isn't she?
This is delicious, isn't it?
Lovely day, isn't it?
In earlier work, I have tried to account for the apparently different
functions of different tag questions by assigning to them slightly
different types of explications, and by phrasing some of these explica-
tions in terms of what one knows or doesn't know, and others, in
terms of what one thinks. But further consideration of a wide range of
examples has convinced me that the distinction between tag questions
seeking factual confirmation and tag questions checking sameness of
opinion was not always feasible. Consider, for example, the following
exchange:
'Mr. Randolph got enough books to start a public library, haven't
you, Mr. Randolph?'
'Well, I do have a few', he chuckled. (GH)
In this case, the tag question refers to a factual matter, not to a matter
of opinion; but what the speaker is seeking is not information or
verification. Tag questions of this kind can therefore be represented
neither in terms of 'I don't know - you may know', nor in terms of 'I
think this - I want to know if you think the same'. On the other hand, a
different formula, couched in terms of 'saying' rather than 'knowing' or
'thinking', can accommodate all confirmation-seeking tag questions. I
suggest, therefore, an explication along the following lines:
S, [opposite-polarity tag]?
(e.g. Maria is Italian, isn't she?
Maria is lovely, isn't she?
Mr. Randolph got enough books to start a public library,
haven't you, Mr. Randolph?)
I say: S
[Maria is Italian; Maria is very nice; Mr. Randolph got
enough books to start a public library]
I think you would say the same
226 The semantics of illocutionary forces
I know: you may not want to say the same
I want you to say if you would say the same
I think you will say you would say the same
As for matching-polarity tags, such as:
Sally is pregnant, is she?
You have bought a house, have you?
Cattell (1973) has shown that they usually echo the interlocutor's earlier
utterances. For example, if John asks Harry about the meaning of a
Russian sentence and Harry supplies that meaning, John could echo
Harry's explanation saying:
It means 'necessity is the mother of invention', does it?
He could not respond by saying:
It means 'necessity is the mother of invention', doesn't it?
I would add, however, that matching polarity tags can also be used as an
interpretation, rather than as an echo of the interlocutor's utterance, as in
the following exchanges:
Hannah: I know, but couldn't you just accept it as security for
a few days' stay here?
Maxine: You're completely broke, are you? (NI)
Shannon: They can't go back in toooowwwwn! - whew - ... Are
they getting out of the bus?
Maxine: You're going to pieces, are you? (NI)
One can account for both the echoing and the interpreting function
of matching polarity tags by means of the following formula:
S, [matching-polarity tag]?
I say: S (e.g. you are completely broke)
I say this because of what I see/hear
I don't want to say that I know this
I want you to say that I can say I know this
The speaker is not really checking whether she has heard correctly
(often, this would be impossible to doubt). Rather, she is checking her
interpretation of what she has heard - or pretending, for whatever
reason, that she is checking it (to be sarcastic, to savour a piece of
Tag questions 227
news, to give herself a chance to reflect on what she has just heard, to
encourage the interlocutor to elaborate, and so on).
In fact, it seems that although matching-polarity tags are typically
used in response to speech, they can also be used in response to
something that one can see looking at the addressee. To allow for this
possibility, the second component in the explication has been phrased
as 'I say this because of what I see/hear' rather than 'I say this because
of what you say'.
5.2. Tags with imperative sentences
Turning now to tag questions added to an imperative, I should like to
point out a wide range of illocutionary forces encoded in such struc-
tures - illocutionary forces which are fully 'determined' and which,
nonetheless, cannot be expressed by means of simple performative
verbs. Compare, for example, the following three constructions:
a. Sit down, will you?
b. Sit down, won't you?
c. Sit down, can't you?
These three patterns are of course closely related to the three 'interroga-
tive' patterns discussed earlier:
a. Will you sit down?
b. Won't you sit down?
c. Can't you sit down?
But despite this close relationship, the imperative patterns cannot be
fully reduced to the interrogative ones (or vice versa) and they merit a
separate discussion, and a separate set of explications.
In some circumstances, the three imperative constructions in question
(will you?, won't you?, and can't you?) may seem to be interchangeable,
but the force of each is quite distinct. In fact, while it is easy to imagine
situations where either (a) and (b) or (a) and (c) are interchangeable, it is
difficult to do so with respect to (b) and (c). Wishing to be very polite to
a distinguished visitor, one would probably say (b) rather than (a), and
one would certainly avoid (c). (c) implies that the addressee should have
already performed the action, that it is a bad thing s/he didn't, and that
the speaker feels some impatience or even irritation because of that.
228 The semantics of illocutionary forces
(b) and (a) are close to each other, but they differ in their implications
concerning the beneficiary. Won't you implies that the action is seen as
something that the addressee can be expected to want to do, whereas will
you implies that it is seen as something wanted by the speaker. For
example (as pointed out to me by Emmie O'Neail), if the addressee is
pacing restlessly up and down the room, getting on the speaker's nerves,
the speaker will be more likely to say Sit down, will you rather than Sit
down, won't you. Typically, won't you is used in situations when the
action is seen as beneficial for the addressee, not for the speaker. Since,
however, won't you can also be used in utterances such as:
Give me a hand, won't you?
I have formulated the relevant component as 'I think you might want
to do it', rather than 'I think that it would be good for you to do it'.
But of course will you, unlike can't you, doesn't always imply
negative feelings. It is often used simply to convey a request. It can
also be used to convey instructions and directions. Won't you, on the
other hand, is frequently used in offers and in invitations. For example:
Pass me a piece of toast, will you, Gilly? (GH)
Make the beds, will you? (GH)
Sit down for a minute, won't you? (GH)
Come back for another little visit, won't you? (GH)
This association of will you with requests and of won't you with offers
tallies very well with the idea that will you presents the action as
desirable from the speaker's point of view, whereas won't you presents
it as something that the speaker thinks the addressee would want, or
would be willing to do.
I would add that the combination of an imperative with will you
constitutes a fairly confident request and that it is often used in
asymmetrical relationships. The confident, self-assured nature of this
frame can be accounted for by the component 'I think you will do it'.
The fact that the combination of an imperative with won't you sounds
more tentative, can be accounted for by the component 'I don't know
if you will do it'.
Sit down, will you?
I say: I want you to do X (sit down)
I don't know if you will do it
Tag questions 229
I think you will do it
I say this because I want you to say that you will do it, and do it
Sit down, won't you?
I say: I want you to do X if you want to do it
I don't know if you will do it
I think you might want to do it
I say this because I want you to say if you want to do it and do it
if you want to do it
Sit down, can't you?
I say: I want you to do X
I see you are not doing it
I feel something bad because of this
I want you to say if you can't do it
I think you can't say it (that you can't do it)
I think you can do X and don't want to do it
I think you should know that this is bad
I say this because I want you to feel something bad and to do X
because of this
5.3. Why can't you (do X)!
Sentences such as
Why can't you understand and be generous - be just! (WE)
are very close in force to imperatives followed by Why can't you:
Why can't you leave me alone!
Leave me alone, why can't you!
It might even be suggested that the two types may have the same
force, and that they may be derived from the same underlying structure
(cf. Sadock 1974).
But in fact the two constructions, though fairly close, are distinct,
and must be defined differently. The imperative construction is restricted
to the second person, and it constitutes an attempt to influence the
addressee. The interrogative construction allows third-person as well as
second-person targets, and since we can't influence by speech someone
who doesn't hear us, this construction can't constitute an attempt to
influence other people. For example, the sentences:
230 The semantics of illocutionary forces
Why can't the English learn how to speak!
Why can't a woman be more like a man!
are not attempts to amend the ways of the English, or of women. Both
constructions state a certain negative perception (W is not doing X),
both express a critical judgment which the speaker thinks should be
shared by the culprits (I think W should know that this is bad), both
attribute the culprits' failure to do what they should be doing to ill-will,
not to inability (I think that W can do X and doesn't want to do it), and
both appeal, rhetorically, for an explanation, an explanation which, in the
speaker's view, can't be forthcoming (I want someone to say, if they
can, why W can't do it; I assume nobody could).
But the imperative construction expresses the speaker's will to change
the situation, as well as a desire to make the addressee 'feel bad' (I say
this because I want you to feel something bad and to do X because of
this). In the interrogative construction, the illocutionary purpose seems
to consist merely in expressing one's thoughts. Moreover, the imperative
construction conveys a current perception (I see that W is not doing X).
By contrast, the interrogative construction states an opinion, which
could easily be a 'repeat'. For example, it seems likely that a person
who says Why can't the English learn how to speak! once, would say
it many times. It is appropriate, therefore, to phrase the relevant com-
ponent of the interrogative construction as 'I say: W is not doing
something that W should be doing'. This tallies well with the general
impression that Why can't they do X is primarily a kind of criticism,
irritated but powerless, whereas Do X, why can't you! is primarily a
kind of directive, aimed at correcting an unsatisfactory situation.
Do X, why can't you!
I say: I want you to do X
I see you are not doing X
I feel something bad because of this
I think you should know this is bad
I want you to say, if you can, why you can't do X
I think you can't say why
I think you can do X and don't want to do it
I say this because I want you to feel something bad and to do X
because of this
Why can't W do X!
I think: W is not doing what W should be doing (X)
Tag questions 231
I feel something bad because of this
I think W should know this is bad
I say: I want somebody to say, if they can, why W can't do it
I think nobody can say it
I think W can do X and doesn't want to do it
I say this because I want to say what I think about Wand what I
feel because of it
5.4. OK?
OK? is an ubiquitous English tag, and yet little or no attention has
been given to it in the literature on the English language. When it
accompanies an imperative its value may seem to be fairly close to that
of will you:
Wait a minute, will you?
Wait a minute, OK?
But in fact, OK? can't always be substituted for will you, as the follow-
ing contrast in acceptability shows:
Shut up, will you (will-ya)?
?Shut up, OK?
There is, so to speak, more to OK? than to will you? Will you is check-
ing merely the addressee's willingness to do something, whereas OK? is
also appealing to the addressee's judgement, and is soliciting something
like agreement, as well as an indication of willingness to comply. There
is also an implication that the speaker's judgement, as well as will, is
involved. For example:
Gilly, give Maime Trotter half a chance, OK? (GH)
You need anything, honey, just let Trotter know, OK? (GH)
In both cases, the speaker proposes a course of action as something
'good', as well as 'wanted', and is checking the speaker's understanding
and acceptance of what is being proposed.
Consider also the following example:
Then she thought better of it. 'You do it, William Ernest, OK?'
(GH)
232 The semantics of illocutionary forces
Here, will you? could not be substituted for 'OK?' - presumably, be-
cause the contrastive stress on you shows that the addressee is being
invited to consider a new variant of the situation. Will you?, it seems,
can only be directed at the addressee's willingness, not at his or her
thought processes.
Most importantly, OK?, in contrast to will you? can also be used with
declarative sentences (with a subject other than 'you'), for example:
I'll stop by for you every day, OK? (*will you) (GH)
The condition seems to be that future or future/present tense, not past
tense, is used:
*1 stopped by for you every day, OK?
I'm leaving it here, OK?
To account for all these features of OK? I would propose the follow-
ing explication:
OK?
I say: I want this to happen
I think it will be a good thing
I know: it will not happen if you don't want it
I want you to say if you want it
I say this because I want it to happen
6. Personal abuse or praise: You X!
Utterances such as
You filthy swine!
You fool! You dickhead! You idiot!
are so common in English that a linguistic theory which would be unable
to analyse them and which would consequently rule them out as 'un-
grammatical' would, I think, have to be regarded as a curiosum.
Utterances of this kind convey with perfect clarity an identifiable
illocutionary force, or rather, three alternative illocutionary forces,
depending on the semantic class to which 'X' belongs.
Personal abuse or praise: You XI 233
a. You liar!
b. You angel! You darling!
c. You beauty!
To see that these three different types cannot be collapsed under one
broader category, consider the following regarded by most speakers as
unacceptable:
*you student!
*you killer!
?You wog!
Types (b) and (c) can be said to constitute minor types, but (a) is a
major category of utterances, with a wide range of 'X's' and a high
frequency of use. 'X' (in category a) has to fall into a general category
of names of person which describe a person as someone who habitually
does something bad and which convey a negative feeling. Student does
not qualify because it doesn't refer to any activity regarded as bad,
killer because it has no negative feeling built into it, and wog because
it doesn't identify any particular vice. Used with an appropriate modifier,
however, wog becomes perfectly acceptable:
You dirty wog!
You stupid wog!
The full illocutionary force of this category (a) can be spelt out as
follows:
You X [neg.] !
I see that you did something very bad
I feel something bad towards you because of this
I want to say something bad about you because of this
I say:
you are a bad kind of person
one has to feel something bad towards someone like you
I say this because I want to say what I think about it and what I
feel because of it
I think you should feel something bad because of this
Expressions such as you fool! or you monster! can of course be used
jocularly and affectionately (I love you, you fool! I miss you, you little
monster!), but such uses exploit the inherently pejorative character of
words such as fool or monster.
234 The semantics of illocutionary forces
The positive subtype of You X! utterances (b) is not entirely symmetri-
cal with respect to the negative one, as can be seen from the fact that
in many speakers' judgement words of high praise, such as saviour or
saint, are not fully felicitous in it. What is fully felicitous is words
which combine unspecified and hyperbolic praise with affection. An
utterance such as You angel! implies that the addressee has done
something good for the speaker, not something good in general, but
specifically, good for the speaker. As a result, the speaker wants to say
something good about the addressee - and is 'lost for words' ('I can't
think of something good enough to say about you'). Consequently, no
specific good qualities can be mentioned, and the praise intended can
be phrased only in terms of general 'lovability' ('you are a person
towards whom one must feel good feelings'). A claim of 'lovability'
must be distinguished from a purely subjective expression of individual
affection, such as love or sweetheart. One can't say:
*you love! (cf. You angel!)
Love used as a term of address means 'my love'. It expresses a relation-
ship between the speaker and the addressee, and does not describe any
generallovability, as angel and darling do.
The illocutionary force of the you X subtype can be spelt out as
follows:
You X [pos.]!
I see that you did something very good for me
I feel something good towards you because of this
I wouldn't have thought anyone would do this
I want to say something good about you because of this
I don't know what I can say
I say:
you are more than good
one has to feel something good towards you
I say this because I want to say what I think about it and what I
feel because of it
[I think you should feel something good because of this]
Finally, subtype (c) is confined to the expression You beauty! and its
variants such as You little beauty!, You bloody beauty! and You beaut! In
this subtype, you doesn't have to refer to a person but it has to refer to
an agent or an action which delights the speaker and surpasses his/her
expectations. For example, one couldn't say You beaut! or You bloody
Illocutionary forces of grammatical and other categories 235
beauty! at the sight of a beautiful sunset, although one could say it at the
sight of a beautiful horse, or at the sight of a beautiful jump performed
by a horse. In this subtype, too, the speaker can't find words good
enough to describe what s/he sees, and s/he feels 'something good' be-
cause of it; but this time, the 'good feeling' doesn't have to be directed
towards the agent ('I feel something good towards you because of this ').
Furthermore, the exclamation You beaut! is not addressed to the agent.
For example, it can well be uttered by someone watching a football
match on television. It is not expected, therefore, to have any impact on
the agent. The other two sub-types on the other hand, do seem to include
a component aimed at influencing the addressee ('I think you should
feel something bad/good because of this').
The semantic formula for You beauty! and its variants reads:
You beauty!
1 see that something good happened because someone did some-
thing
I feel something good because of this
I wouldn't have thought I would see this
1 want to say something good about it because of this
I don't know what I can say
I say:
this is beautiful
one has to feel something good because of this
I say this because I want to say what I feel because of it
7. Illocutionary forces of grammatical and
other categories
7.1. Modal verbs
Another category of clues to illocutionary force comprises modal
verbs, used as in the following exchanges (whose dialectal aspect is
irrelevant in the present context):
Marthy: How old' II she be now?
Chris: She must be - lat me see - she must be twenty year ole,
py Yo! (AC)
236 The semantics of illocutionary forces
The combination of the modal will with the adverb now shows
clearly that in this sentence will does not indicate future tense. What
it does indicate, in this particular sentence, is that the speaker is
inviting the addressee to make an estimate. Most generally, it indicates
a thinking process, leading a person from not being able to say the right
thing about something to being able to do so, with some confidence:
Johnny: Where's it from?
Larry (after a glance): St. Paul. That'll be in Minnesota, I'm
thinkin'. (Ae)
Quite clearly, the speaker is not informing the addressee that 8t. Paul is
in Minnesota, and he is not exactly inferring it either. There is no perfor-
mative verb in English which would capture exactly the illocutionary
force of such sentences. Nonetheless, its force can be clearly stated:
X will be P
I don't say I know this [X is P]
I think I'll know it if I think about it
I want to think about it
I think I know it now
I say: [ ]
A question such as How old'll she be now? invites an answer based on
current thinking effort:
I think you will know it if you think about it
I want you to think about it
The explication has temporal order built into it, referring as it does to
a passage of time. It shows that the modal will is related to the future
will, and that in fact the meaning of the latter is contained in the meaning
of the former. (Consequently, the explication is not circular, as it reduces
the modal will to the temporal will. It is not my purpose to explicate the
latter in the present context.)
As for the modal must (She must be at least forty), it indicates a
similar, but not identical illocutionary force. It is less speculative, and it
doesn't refer to time spent in thinking. Looking at the portrait of an
unknown woman, one would say:
She must be very beautiful.
Illocutionary forces of grammatical and other categories 237
but hardly
?She will be very beautiful.
Since in this particular situation the judgment is clearly based on what
one can see (the portrait), not on a process of thinking, will is inappropri-
ate. Of course the sentence is fine if will is interpreted in the future
sense 'she will be beautiful in the future', but not if it is interpreted in
the modal sense 'she will be beautiful now'.
The illocutionary force of must sentences can be stated as follows:
X must be P
I say: I think I can say this about X
I don't say I know it
I think one must think this
I say this because I want to say what I think
This is of course a much more tentative force than that of / must say
sentences. The modal must in sentences of the X must be P type is not
derived from an underlying 'I must say' component. Rather, it is derived
from 'I think one must think this'. In saying 'X must be P' the speaker
stops short of asserting that X is P and the explication doesn't include
a component of the form 'I say: X is P'. But an explication of the frame
'I must say: X is P' would have to include such a component.
For example, if the speaker is telling us a story about some unknown
woman, we could respond by saying 'She must be very beautiful', but
not by saying 'I must say, she is very beautiful' - because the latter
sentence would imply a personal assessment of the woman's beauty,
which is impossible in the case of someone unknown. Compare also
the following two sentences:
This must be the milkman.
?/ must say, this is the milkman.
The first sentence expresses a rather confident guess and it is of course
perfectly acceptable; but the second sentence is odd, because it suggests
a personal judgment concerning a non-factual matter, whereas the propo-
sition introduced in this frame is in fact factual. The personal judgment
in question appears to be expressed reluctantly but without the slightest
hesitation as to its validity. This can be portrayed as follows:
I must say, she is very beautiful (?this is the milkman)
I say: she is very beautiful
238 The semantics of illocutionary forces
I don't say it because I want to say it
I must say it if I want to say what I think
I say it because I want to say what I think
The non-factual character of the frame 'I must say' is accounted for by
the combination of a confident assertion (I say: she is very beautiful)
with an indication of a subjective basis of this assertion ('I say this
because I want to say what I think') and by the absence of the disclaimer
'I don't say I know it'.
7.2. Mental verbs
The illocutionary force of an utterance can be identified quite clearly by
a 'mental' verb used in it (present tense, first person singular), which
identifies the speaker's mental state associated with that utterance.
(Cf. Urmson 1963; Ross 1973.) A sentence introduced by a frame such
as I think, I hope, I fear, I reckon, I suppose, I gather, or I remember,
without a complementiser, makes it quite clear that the speaker is
not informing the addressee, not announcing anything, and not even
telling the addressee anything, but that s/he is rather expressing his/
her thoughts.
You have a young lady, I understand. (GP)
I gather Donald has a girl too, is that right? (GP)
Your wife's working, I gather. (GP)
Thus, the illocutionary purpose of an utterance introduced in a first-
person mental frame (without a complementiser) seems to be:
I say this because I want to say what I think
Given this illocutionary purpose, it is perhaps not necessary to include in
the explication an explicit disclaimer: 'I say this not because I want other
people to think this'. If the speaker shows clearly that what s/he wants
is to express his/her thoughts, then it is more or less implied that s/he is
not motivated by a desire to influence other people.
It seems, however, that a mental illocutionary clause serves as a dis-
claimer in a different sense, and this other disclaiming function should
perhaps be portrayed explicitly in the semantic formula. A person who
says:
Illocutionary forces of grammatical and other categories 239
I believe he's quite a decent fellow.
I hope she comes.
indicates that s/he does not want to commit himself/herself to what s/he
says. S/he indicates that s/he is not prepared to say (to 'assert') the
proposition in question. For this reason, if an utterance such as He's
quite a decent fellow can be represented as:
I say: he's quite a decent fellow
and is compatible with the illocutionary purpose:
I say this because I want you to know it
an utterance such as: I believe he's quite a decent fellow should, I think,
be represented as:
I say: I believe this [he's quite a decent fellow]
I don't want to say that I know this
I say this because I want to say what I think
The sentence in the I believe frame would not be reported as, for
example:
She claimed that he was quite a decent fellow.
It could only be reported as:
She said she thought (believed) he was quite a decent fellow.
This mode of reporting confirms the supposition that in sentences in the
frame I believe, the complement of I believe is not embedded directly
under I say, and that I believe is so embedded.
I offer here, for comparison, explications of the I fear and I gather
frames:
S, I fear (e.g. Too late now, Don, I fear)
I say:
I think this [it is too late now]
this is bad
I don't want to say I know this
I say this because I want to say what I think
S, I gather (e.g. His wife's working, I gather)
I say: I think this [his wife is working]
I don't want to say I know this
240 The semantics of illocutionary forces
I think this because of what I hear/see
I say this because I want to say what I think
(In the last example I have replaced you and your with he and his,
because you frames have their own peculiarities, which have to be
discussed separately.)
7.3. Particles and conjunctions
As the ancient and medieval theorists of language well realised, the
illocutionary force of an utterance is often signalled by particles, con-
junctions and interjections. For example, the Greek particle hos 'how'
was regarded by the Stoics as a marker of what they called thaumastikon,
the lekton of admiration or astonishment (see Nuchelmans 1973:63; see
also note 4), and the Latin particle utinam 'would it' was described by
Paul of Venice as nota optandi, the mark of a wish (Nuchelmans
1973:48). In a similar vein, Austin suggested that "we may use the
particle 'still' with the force of 'I insist that'; we use 'therefore' with the
force of 'I conclude that'; we use 'although' with the force 'I concede
that'." (Austin 1962:75).
Clearly, there is also a close relationship between yes and I agree,
between no and I disagree, or between and and I add. Nonetheless, no
particle or conjunction has exactly the same meaning as a performative
verb. In fact, particles and conjunctions usually specify only some parts
of the illocutionary force, whereas a performative verb specifies all of
it. Consequently, particles and conjunctions, in contrast to performative
verbs, can be combined with a fairly wide range of other devices
partially specifying the illocutionary force. Thus, one can say:
So you've just landed? (AC)
It'll get him all right, then? (AC)
using so or then in the context of a question, but one could hardly use a
phrase such as I infer or I conclude in this way.
In this section I will discuss these two words, so and then, in some
detail, to show how the proposed method of analysis is applied to
discourse particles and conjunctions of this kind. (Cf. also Chapter 9
below on particles.)
Halliday - Hasan (1976:257) suggest that when so and then are used
as causal conjunctions they indicate "the speaker's reasoning process: 'I
Il.locutionary forces of grammatical and other categories 241
conclude from what you say (or other evidence)' - compare expressions
such as I gather." I think that this explication is basically correct, but
that it has to be refined. For one thing, so is not identical in meaning
with then, so two slightly different formulae are needed rather than one.
For another, the verb conclude is too intellectual to portray the impact
of so. Concluding means not so much saying something on the basis
of the evidence available, as saying something on the basis of an
examination of the evidence. For this reason, concluding is based on
arguments, on reasoning, not directly on evidence. But so refers directly
to the evidence perceived.
For example, in the first scene of 'The night of the iguana' (Williams
1961) one of the protagonists, Shannon, is told by another, Maxine,
that the latter's husband, Fred, is dead. A few moments later Shannon
resumes the topic, saying:
So Fred is dead?
Clearly, Shannon is not concluding that Fred is dead, because she has
been informed of it. Her attitude can't be portrayed, therefore, as: 'I
conclude from what you say that Fred is dead'. It can be portrayed,
however, as follows:
I think I can now say I know 'Fred is dead'
I think this because of what you say
If so is combined with the interrogative intonation, as in the example
under discussion, another component is signalled in addition:
I want you to say something more about it
and possibly even one more:
I want us to talk about it (i.e. to say some things about it to one
another)
Thus, a full explication could perhaps read as follows:
So S? (e.g. So Fred is dead?)
I think I can now say I know S because of what you say
I want you to say that I can say it
I want you to say something more about it
I want us to say some things about it to one another
242 The semantics of illocutionary forces
Then differs from so in a number of respects. Unlike so, it doesn't
refer to the immediate, audible evidence. For this reason, it can be used
in more hypothetical contexts, where so is out of place, as in the follow-
ing examples repeated here after Halliday - Hasan (1976:258):
'And what does it live on?'
'Weak tea with cream in it.'
A new difficulty came into Alice's head.
'Supposing it couldn't find any?' she suggested.
'Then [*so] it would die, of course.'
Then can also be used in making a judgement, i.e. in assessing a
situation.
'Have some wine', the March Hare said in an encouraging tone.
Alice looked all around the table, but there was nothing on it but
tea. '[ don't see any wine' , she remarked.
'There isn't any', said the March Hare.
'Then [?so] it wasn't very civil of you to offer it', said Alice
angrily.
All these differences in usage between the causal so and the causal then
can be predicted, I think, from the following two explications:
so: I think I can now say I know this because of what you say
then: I say this because of what you said now
So, unlike then, is indeed closely related to the expression [ gather, and it
wouldn't make much sense to use [ gather in expressing an evaluation:
?[ gather that it wasn't very civil of you to offer it.
Both so and [ gather indicate that the speaker is trying to say a true
sentence on the basis of what s/he perceives; by contrast, then indicates
that the speaker is saying something because of what has been said
before.
But the semantic differences between so and then can't be equated
with the differences between gather, conclude, infer, or other illocution-
ary verbs. For example, gather, though closer to so than it is to then,
is less confident and less directly linked to evidence than so is. Having
heard a few minutes earlier, from Maxine, that Fred is dead, one couldn't
say to Maxine:
[ gather that Fred is dead.
Illocutionary forces of grammatical and other categories 243
Each verb, and each particle, has to be described separately, and though
many components recur, each verb and each particle constitutes a unique
configuration of components. (See Goddard 1979; Wierzbicka 1976.)
7.4. Interjections
Interjections express a feeling or a 'want' on the speaker's part. They
have, therefore, their own illocutionary force, which can be described in
terms of components such as 'I feel X', 'I want Y'. Since, however, they
typically combine with other utterances into larger wholes, and since
their illocutionary force must be compatible with that of the co-utter-
ance, they often serve as important clues identifying the illocutionary
force of the combined utterance as a whole. (A more comprehensive
study of interjections is contained in Chapter 8 below.) For example, in
the utterance:
Ah, my God, are they still in the bus? (NI)
the part are they still in the bus? will not be interpreted as a pure
question, despite its interrogative structure, because the meaning of the
interjection Ah (or Oh) my God ensures that what follows is to be inter-
preted as a sudden realisation:
Oh my God, Sf
1 think 1 now know something bad is happening (S)
1 wouldn't have thought this would happen
1 feel something bad because of this
1 say this in this way because of this
The related interjection for God's sake encodes a slightly different
attitude: it implies that although something bad is going on, the speaker
intends to stop it; and that as s/he is not in control of the situation,
s/he will attempt to stop it via an action of the addressee. Hence the co-
utterance will probably include an imperative, and, moreover, a bare
imperative, unembellished by any interrogative devices, such as tag
questions, why-don't-you's, etc.
The illocutionary force of an utterance in the For God's sake frame
can be spelt out as follows:
For God's sake, Sf
1 see something bad in happening
244 The semantics of illocutionary forces
I feel something bad because of this
I don't want this to happen
I want you to do something
I want you to do it now
I say this in this way because I can't do anything else
I think that you have to do it because of this
Using a more conventional (but less precise) language, one might say
that for God's sake signals a negative judgment, negative feelings,
frustration, helplessness, a sense of urgency, an appeal to the addressee,
a feeling of not being in control, combined with a desire to obtain
one's goal no matter what, a sense of desperation.
As a final example, let us consider the interjection gee. Often
described in dictionaries as an interjection of surprise, gee has in fact
a more complex meaning; and 'surprise' is sometimes an inappropriate
gloss, as well as an inaccurate one:
Gee, that's a nice dress!
Gee, I was scared for a moment I killed you! (AC)
Gee, you look like you had it! (NI)
Gee, wasn't I sick of it - and of them! (AC)
The first example expresses appreciation or admiration. In the second
the speaker re-lives a recently experienced fear. The third expresses
a mild pity. In the last, the speaker re-lives a past disgust. It is not
surprise, then, which constitutes the semantic invariant of gee. The real
invariant, however, is related to surprise in so far as it combines an
emotional component with the idea of unexpectedness:
Gee!
I am thinking about X
I feel something because of this
I wouldn't have thought I would see (feel) this
It is easy to check that this set of components does fit all the examples
of utterances with gee cited above.
Generally speaking, it seems that while many interjections and other
illocutionary devices encode an emotion, the nature of this emotion is
never very specific. If we tried to include in the semantic representation
a name of the emotion (such as surprise, anger, irritation, frustration,
etc.) we would be overspecifying the emotion conventionally conveyed
- just as we would be over-specifying the illocutionary force of a
Illocutionary forces of grammatical and other categories 245
syntactic construction by reconstructing a speech act verb). The level of
specificity of emotions encoded in various illocutionary devices doesn't
seem to go beyond 'I feel something', or 'I feel something bad/good' or
'I feel something bad/good towards you'. Additional information about
the kind of emotion involved is conveyed implicitly, and neither can
nor should be specified in the semantic representation.
Returning to gee, I would add that besides having its own illocution-
ary force, it functions also as an important clue to the illocutionary force
of the co-utterance. The point is that an utterance framed by gee is not
used for informing, stating facts, reporting, reminding, warning, and
countless other purposes which can be served by declarative sentences.
Gee signals that the force of the co-utterance is this:
I say this because I now think this
In fact, gee shares this function with many other interjections. For
example, the sentence:
I left the oven on.
could be reported as 'she informed him that ', 'she confessed that ... ',
'she admitted that ... ', 'she stated that ', and so on. But when
framed by Oh, my God, it could not be so reported.
Oh my God! I left the oven on!
In this frame, the co-utterance expresses a spontaneous thought:
I say this because I now think this
7.5. Fixed expressions
In addition to lexical devices such as conjunctions, particles and interjec-
tions, and to syntactic constructions, English (and presumably every
language) contains a large number of fixed expressions, encoding a
variety of illocutionary forces. In fact no sharp line divides these 'fixed
expressions' from productive lexical and grammatical resources. For
example, expressions such as Oh, my God! or For God's sake! can be
described either as 'fixed expressions' or as 'complex interjections', and
both descriptions would be equally apt.
Similarly, expressions such as Why don't you or How about can be
described either as markers of specific syntactic constructions or as
246 The semantics of illocutionary forces
'fixed expressions'. Since all 'fixed expressions' have their 'syntactic
fields' (i.e. more or less limited spheres of use), they can be seen as
pertaining to syntax as well as to the lexicon.
The clearest examples of fixed expressions carrying their own
illocutionary force come from the area of greetings and leave taking.
Roughly speaking, one can describe the force of expressions such as
Good morning or Good afternoon in terms of the verb greet. But no
similar verb exists in contemporary English which would correspond
to the expressions Good-bye or Good night. The illocutionary force
of the latter expressions, then, can be stated only in terms of a set of
components. For Good-bye, this can be done as follows:
Good-bye
I think you and I both know:
after now we will not be in the same place
we will not be able to say things to one another (any more)
because of this
I want to say something to you because of this
I want to say something of the kind that people say to one
another at a time like this
I say this because I know I should say something like this
Expressions such as Good morning or Good night are of course
very frequently used, and it is hard to overlook them in an over-all
description of language use. But English, and presumably other
languages as well, have countless other expressions, which while less
frequently used, are equally clear in their illocutionary force. I will
illustrate this claim with two examples: How dare you! and Go
(and) jump in the lake! For reasons of space, I won't discuss these
expressions, but I will simply propose semantic formulae to represent
their illocutionary force.
How dare you!
I see (perceive) this: you are doing something bad
I don't want you to do this
I wouldn't have thought you would do it
I say: I don't know 'how you can dare' to do this
I feel something bad because of this
I say this because I want you to know what I think and
what I feel because of this
Illocutionary forces of grammatical and other categories 247
Go (and) jump in the lake!
I know that you want me to do something
I think you know that I don't want to do it
I will not do it
I think you want me to think:
if I don't do it something bad can happen to you
I say: I don't want to think about this
I say:
if something bad happens to you
I will not think/feel something bad because of this
I say this because I don't want to have to say more about it
I will confine myself to just two comments on these explications. It
will be noticed that the surface form of the expression How dare you!
incorporates fragments of the proposed underlying form, whereas the
expression Go and jump in the lake! does not. This means, that, for
example, the lexical item dare occurs in the fixed expression under
consideration in the same meaning which it carries elsewhere: the fixed
expression adds additional components to the meaning encoded by its
constituents, but the usual meaning of those constituents is preserved. On
the other hand, the expression Go and jump in the lake! can be said to
be more deeply 'idiomatic', because the usual meaning of its lexical
constituents is not preserved here; accordingly, words such as jump or
lake are not mentioned in the proposed explication at all.
7.6. Intonation
It goes without saying that no survey of the illocutionary devices of
natural language could be complete if it failed to mention intonation.
The present chapter has no ambition of providing a complete survey (a
solid book would be necessary for that). Nonetheless, I feel that even
here intonation must be at least mentioned.
It seems indubitable that intonation plays a fundamental role in the
area of illocutionary meanings. Limitations of space (as well as of my
competence) preclude any serious discussion of this problem here. I
would like, however, to illustrate this role of intonation with a short
quote from what seems to me to be a highly competent and illuminating
source of information on the subject (Deakin 1981). With respect to
'positive interrogative exclamations', Deakin says:
248 The semantics of illocutionary forces
Utterances like 'is she beautiful' with rhythmically repeated rising-falling
pitch glides and lengthening on each syllable (especially the last) are not
taken as meaning an inquiry, or even that the addressee is expected to
reply. Rhetorical questions like 'is this the way we should be doing things'
with 'high' pitch and 'tense' voicing are not taken as inquiries, or as
expecting an answer (although the answer 'no' is consistent with them).
Similarly, 'can you pass the salt' uttered with low rising intonation is
generally taken as a request, not an inquiry. The addressee does not have
to reply, unless he does not comply with the request. He can readily
comply and say nothing, although it is more usual to reply as well.
(Deakin 1981:57)
It is true that the best authorities on intonation, such as Bolinger
(1982), caution against too hasty attempts to link specific intonational
features with specific kinds of speech acts. But while Bolinger's
strictures are probably fully justified with respect to his particular
targets, cautious and careful semantic analysis of intonation of the kind
attempted by Deakin seems to me not only justifiable but urgently
needed. Just as the illocutionary force conveyed by grammatical and
lexical means can't be adequately portrayed in global terms (for ex-
ample, by means of English speech act verbs), neither can the illocution-
ary force conveyed by means of intonation. Arguably, however, it can
be portrayed by means of more subtle analytical tools. Deakin (1981)
goes so far as to postulate specific illocutionary meanings for the basic
English tones, and for the falling and rising heads, using for that purpose
the natural semantic metalanguage based on semantic primitives. His
analysis seems to me illuminating and plausible.
8. Comparing illocutionary forces across languages
The method of analysis illustrated in this chapter allows us, I think, to
compare illocutionary forces encoded in different constructions in a
precise and illuminating way. It allows us also to compare, in a precise
and illuminating way, attitudinal and interactional meanings encoded
in different languages, and correlated with other cultural and social
differences intuitively more obvious perhaps but harder to document.
In this section, I will illustrate this point by comparing two illocution-
Comparing illocutionary forces across languages 249
ary devices allowing the speakers to express anger (or something like
anger) - one in English, and one in Yiddish.
In English - in particular, in Australian English - people often
express their anger by means of the transitive subjectless pattern:
'V you!', for example:
Damn you!
Blast you!
Fuck you!
Of course, the pattern in question does not belong to what is called polite
conversation. It is not, nonetheless, a marginal and negligible feature
of English speech. On the contrary, its role in the universe of English
discourse can be seen as significant.
Formally, this pattern appears to violate the general rule that the 'de-
leted' subject of an imperative is 'you'. In this case, 'you' is the object,
not, as one would expect, the subject, and this reversal contributes to the
expressive force of the expressions in question.
The set of verbs which can occur in this pattern is very restricted.
For example, one cannot say:
*Murder you!
*Strangle you!
The underlying generalisation seems to be this: the speaker not only
wants someone 'to do something bad to the addresse', but he also wants
'to say something bad', that is, to break some verbal taboo. This can be
represented as follows:
V you!
(a) I feel something bad towards you
(b) I want to say something bad because of this
(c) I say: I want someone to do something bad to you
(d) I don't want to say more
Component (a) shows the speaker's ill feeling towards the addressee,
component (b) indicates the impulse to break a taboo, and component (c)
spells out the 'bad action' which the addressee .should undergo; compo-
nent (d) shows that the whole utterance is perceived as short and some-
how dismissive.
It is very interesting to compare this English pattern with the charac-
teristic Yiddish pattern described by Matisoff (1979:68):
250 The semantics of illocutionary forces
Ver geharget!
'Be (lit. become) murdered!
Ver dershtikt!
'Be (lit. become) choked!'
Ver geshvoln!
'Become swollen!'
This Yiddish pattern, too, expresses the speaker's anger (ill feelings)
towards the addressee, and a desire that something bad should happen
to him. However, it seems less 'active' and more 'wishful' (helpless?)
than the corresponding English expressions, in so far as it is not
restricted to transitive action verbs; the speaker wants something bad to
happen to the addressee, but not necessarily something bad to be done to
the addressee; and there is no desire, on the speaker's part, to break a
verbal taboo. We can represent the meaning involved as follows:
Be murdered! Be choked! Be swollen!
(a) I feel something bad towards you
(b) I want to say something because of this
(c) I say: I want something bad to happen to you
It is interesting to note that in Yiddish the pattern in question appears
to be restricted to the 'bad feelings' towards the addressee, whereas the
English pattern can be extended to the third person as well:
Damn him! Stuff him!
In Yiddish, bad feelings towards third person are normally expressed in
a somewhat less violent manner, with the auxiliary verb zol 'let/should'
rather than with an imperative of the verb vern 'become'. For example
(Matisoff 1979:66-68):
Platsn zol er!
'May he explode!'
A fayer zol im ontsindn!
'Maya fire ignite him!'
The 'milder' character of such third person curses is manifested also
in their ability to be used in complex sentences, as parenthetical clauses,
or as main clauses, for example (Matisoff 1979:61-71):
Comparing illocutionary forces across languages 251
Mayn man, zol er geshosn vern, hot nekhtn 6ngevorn a sahk gelt.
'My husband, may he be shot, lost a lot of money yesterday.'
Mayn shviger, klog veys ir, hot a beyze tsung.
'My mother-in-law, maya lament be known to her, has a wicked
tongue.'
A shvarts yor oyf ir, a gantsn tog hot zi mir gehakt a tshaynik
vegn kleynikaytn!
'A black year on her, all day long she chewed my ear off with
trivia. '
OJ, zol im nor az6y rinen fun noz, vi's rint mir fun der kvalpen!
'Oh, may his nose only leak on him the way this fountain-pen
leaks on me!'
It is also significant, needless to say, that Yiddish third person curses
are often not only long and flowery but also humourous, for example
(Matisoff 1979:66, 70):
Vos es hot zikh mir gekholemt di nakht un letste nakht, zol zikh
6yslozn tsu dayn kop un layb un lebn!
'May my nightmares of the last two nights let themselves loose
onto your head and body and life!'
A ziser toyt zol er hobn - a trok mit tsuker zol im iberforn!
'May he have a sweet death - run over by a sugar truck!'
For sentences of this kind (third person curses with zol) I would propose
the following explication:
my father-in-law, maya disease enter his gums
(a) I feel something bad towards this person
(b) I want to say something because of this
(c) I say: I would want something bad to happen to this person
The distinction between 'I want' and 'I would want' is not without its
problems, but it captures well, it seems to me, the difference in the
speaker's attitude ('milder' and more tentative in the case of a third
person than in the case of the addressee). In English, however, no similar
distinction is drawn between Damn you! and Damn him!
It is worth noting that the Yiddish curses play a more central role in
Yiddish discourse than the English pattern discussed at the outset. The
English 'Verb you!' pattern is based on a conscious violation of a
252 The semantics of illocutionary forces
taboo; its role in English discourse is therefore, naturally, restricted. The
same can be said about other forms of verbal abuse available in
English. Traditional Jewish culture, however, allows much more free
expression of 'bad feelings' directed at other people; and the language
offers suitable devices for doing so, devices which are more generally
available, and which are compatible with a wider range of genres
and registers.
The use of a language-independent semantic metalanguage based on
natural language enables us to describe the illocutionary forces encoded
in different languages in considerable detail, and to compare them -
both intra- and inter-linguistically - in a precise and verifiable manner.
Precise and verifiable linguistic comparisons of this kind can constitute,
in tum, a reliable basis for cultural comparisons, which in the past were
often based on speculations or on ad hoc, unsystematic, observations
on linguistic differences.
9. Conclusion
I hope to have shown that English and, presumably, any other language,
possesses a whole range of devices which convey well defined illocu-
tionary forces. Since these devices are largely language-specific, their
force cannot be calculated on the basis of any universal pragmatic
maxims, Gricean or non-Gricean. Needless to say, even leaving aside the
intonation, the devices discussed in this chapter constitute only a small
part of the language's illocutionary resources. They should suffice, how-
ever, to establish the point that the alleged enormous indeterminacy of
illocutionary forces is largely an illusion, born out of an inadequate
analytical model. When instead of trying to squeeze every conceivable
utterance into a pigeon-hole created by a speech act verb we analyse the
illocutionary force of each utterance into individual components, it
emerges that language provides numerous unmistakable illocutionary
clues, which enable the listeners - and the linguists - to identify
illocutionary forces with considerable precision (on a sub-conscious level
in the case of listeners, and on a conscious level in the case of linguists).
(For numerous subtle analyses of illocutionary clues in Russian, see
Paduceva 1985.)
Conclusion 253
To return to the question of 'indeterminacy', I think the answer must
be that when people speak to us we usually can detect what they are
doing, i.e. what acts they are performing - or at least we can do it to
a much higher degree than has been assumed. Illocutionary forces are
not nearly as indeterminate as has been maintained. But the exact level
of their determinacy is an empirical matter which has to be established
on an empirical basis.
A preliminary study of a few samples of dialogue taken from plays,
as well as from transcripts of natural conversation, seems to suggest
that, generally speaking, we know fairly well what the speakers are
doing: the samples examined swarm with clear illocutionary clues,
whose force can be spelt out in rigorous semantic formulae. But to show
that this is the case a separate study is required. The main purpose of
the present chapter consists in proposing a framework within which
illocutionary forces can be analysed at all - with precision and without
arbitrariness.
It has also been a purpose of this chapter to show that the decomposi-
tion of illocutionary forces into illocutionary components offers solu-
tions to problems which have led a number of linguists to abandon
the performative analysis and which have driven them to a desperate
position where they are forced to condemn the bulk of conversational
English as 'ungrammatical'.
The performative analysis, which met the problems of language use
head on and which indeed 'took the bull by the horns' (Bach - Hamish
1982:225), was, in my view, on the right track (and in fact, on the track
indicated by the great medieval thinkers such as Peter Abelard or Roger
Bacon). It is true that in the form in which it was proposed by generative
semanticists this hypothesis encountered insuperable problems. But the
problems encountered by the more recent theories, which once again
dig a gulf between grammar and language use, are even worse.
Decomposition of illocutionary forces allows us to preserve the com-
mon sense position and not to have to condemn conversational English
as 'ungrammatical'. It allows us to account for language use, and to
make sense of grammar in terms of people's communicative needs and
illocutionary intentions. It allows us to relate syntactic surface structures
to meaning, and to validate claims about meaning with observations
about the actual language use.
The problem of 'what people do with words' can of course be studied
from a number of different points of view, not only from a purely
linguistic point of view such as that advocated here. In particular,
254 The semantics of illocutionary forces
conversational analysis of the kind practised by sociologists (see for
example Sudnow 1972; Schenkein 1978; Psathas 1979) has, in my view,
a great deal to offer in this respect. But when sociologists interpret,
for example, English tag questions as 'exit techniques' (cf. Sacks -
Schegloff - Jefferson 1978:30), their analysis is on an entirely different
plane from that proposed in this work. A sociological study of conversa-
tion can't represent an alternative to illocutionary linguistics; rather, the
two must complement each other.
I would also venture to suggest that while the former should be of
great interest to the latter, the logic of their relationship is such that the
former should build on the latter, rather than vice versa. I think that the
hopes of linguists who look to sociology for solutions to linguistic
problems (cf. for example Levinson 1983) are bound to be disappointed.
Chapter 7
Italian reduplication:
its meaning and its cultural significance
In this chapter, I want to look at the relationship between semantics and
pragmatics through the prism of one illocutionary device used in one
particular language. The language is Italian, and the device is that of
'syntactic reduplication'. The problem is conveniently 'simple', and so
allows a reasonably deep probe within the confines of one short chapter.
At the same time, it is sufficiently multifaceted to illustrate a wider range
of fundamental theoretical issues such as meaning versus implicature,
grammar versus rhetoric, autonomous grammar versus an integrated
theory of linguistic communication, the boundary between semantics
and pragmatics, iconicity versus arbitrariness, or the relationship
between culture (in the broad sense) and linguistic structure.
1. Italian reduplication: preliminary discussion
In Italian, it is very common to reduplicate adjectives, adverbs, and
adverbial expressions, for, roughly speaking, expressive purposes. I call
the device in question reduplication rather than repetition, because I see
a functional difference (as well as a similarity) between pauseless ex-
pressions such as adagio adagio 'slowly slowly' and expressions such as
adagio, adagio where a comma signals the presence of a pause. At the
same time, I call this phenomenon 'syntactic reduplication' rather than
just 'reduplication', because the process in question operates on words
rather than on morphemes, so that one can say, for example, not only
adagio adagio, but also adagino adagino (where -in- is a diminutive
suffix). Like many other languages, Italian also has some frozen ex-
pressions based on reduplication, but I say nothing about these here,
since this chapter is concerned only with reduplication as a productive
illocutionary device. Grandgent (1908:32) notes the emergence of what
I call syntactic reduplication in Vulgar Latin: "Repetition for intensive
256 Italian reduplication: its meaning and its cultural significance
effect is not uncommon in late writers: Commodian, malum malum ...
bene bene, bonis bonis, fortis fortis, malus malus, etc."
Typically, reduplications of this kind are translated into English by
means of the 'intensifier' very. For example, the Italian expressions
below would usually be rendered in English by the expressions in the
right-hand column:
bella bella
duro duro
zitto zitto
adagio adagio
in fretta in fretta
'very beautiful'
'very hard'
'very quiet(ly)'
'very slowly'
'very hurriedly'
But the possible range of use of Italian reduplication is much wider
than the range of use of the word very in English. For example, the
expression neri neri ('black black') would probably not be translated
into English as very black. Compare, for example, the following
sentence:
Due occhi, neri neri anch' essi, si fissavano tallora in viso aIle
persone, con un'investigazione superba ... (Manzoni 1972:235)
with its two English translations (by different translators):
Sometimes she would fix two very dark eyes on another's
face with a piercing look of haughty investigation ... (Manzoni
1914, 1:154)
A pair of eyes - jet black, too - would sometimes fasten on
people's faces with an air of haughty curiosity ... (Manzoni
1968: 116)
The first translator does use the word very, but simultaneously substitutes
dark for black; the second translator translates neri neri as jet black.
Even more remarkably, however, syntactic reduplication can be used
in Italian in contexts where no 'qualities', gradable or otherwise, are
spoken of, as in the following examples:
... se no, lascio Ie mie scuse, e me ne vo diritto diritto a casa
mia. (Manzoni 1972:578)
'... if not, [' II leave my excuses, and go straight off back home.'
(Manzoni 1968:323)
Italian reduplication: preliminary discussion 257
Di grazia, ... un po' di luogo, un pochino; appena appena da
poter passare. (Manzoni 1972:344)
'Please, gentlemen, ... a little room, just a very little - just
enough to let us pass.' (Manzoni 1968:183)
se rimaneva Ii' in ginocchio, ancora per qualche momento, quasi
quasi gli chiedevo scusa io, che m' abbia ammazzato il fratello.
(Manzoni 1972: 119)
'If he'd stayed down on his knees a moment longer, I'd almost
have got to the point of asking his forgiveness myself, for
having killed my brother for me.' (Manzoni 1968:52)
. .. e che mi faccia la carita di venir da noi poverette, subito
subito. (Manzoni 1972:95)
, ... and would he do us poor folk the kindness of coming to us
straight away.' (Manzoni 1968:39)
This shows that whatever the function of Italian reduplication may be, it
is not the same as that of 'intensifiers' such as the English word very or
its Italian equivalent molto. To say that bella bella means 'very beauti-
ful' or that leggera leggera means 'very light' is not only inaccurate but
could also be misleading, as it encourages false predictions. It leads one
to predict, for example, that the Italian expressions subito subito, quasi
quasi, appena appena or diritto diritto should be as ungrammatical as
the English expressions 'very at once', 'very almost', 'very barely' ('very
just'), or 'very straight' ('I'll go very straight home') would be. In fact,
however, the Italian expressions in question are perfectly grammatical
and perfectly felicitous (unlike their alleged English counterparts).
Italian grammars usually characterise the function of reduplication
(raddoppiamento) as 'intensification' (l' intensificazione). For example,
Lepschy - Lepschy (1984:103) write: "L'intensificazione di un agget-
tivo si puo ottenere, oltre che combinandolo con un avverbio, anche
attraverso la ripetizione: una stanza molto piccola 0 piccola piccola."
[The intensification of an adjective can also be achieved, in addition
to combining it with an adverb, by repetition: 'a very small room', or
'small small'.] But again, one has to say that this characterisation is
misleading because it suggests that the 'repetition' is interchangeable
with the adverb molto. But one cannot say molto subito 'very at once'
as one says molto piccola 'very small', even though one can say
both subito subito and piccola piccola.
258 Italian reduplication: its meaning and its cultural significance
One could always say, of course, that the function of the Italian
reduplication is that of 'emphasis' - as people often do when speaking
of devices whose exact force they are unable to state. But this would
hardly have much explanatory value, as 'emphasis' is also invoked as
an explanation for a whole range of other devices, in Italian and, it
seems, in every other language which has ever been described (heavy
stress is used 'for emphasis', particles are used 'for emphasis', repetition
is used 'for emphasis', and so on).
As I see it, the reduplication illustrated with expressions such as bella
bella or subito subito is a characteristically Italian illocutionary device,
whose exact function and force is revealed neither by means of rough
translation equivalents from other languages (such as the English word
very) nor by means of vague and opaque labels such as 'emphasis'. It
can only be revealed by means of a semantic formula which would fit
all the contexts where the device in question can be used - and only
those contexts. (When I say 'characteristically Italian' I do not mean
'uniquely Italian', cf. Triandaphyllidis' (1975:653) account of a similar
device in Modem Greek.)
2. Discourse and illocutionary grammar
I would like to put forward a general hypothesis to the effect that 'illocu-
tionary grammar' is born in the ethnography of spoken discourse (see
Hymes 1962). What I mean is this. Every language has its own set of
language-specific illocutionary devices, encoding specific illocutionary
meanings. A set of this kind can be called the 'illocutionary grammar'
of a given language. In addition, there are universal or near-universal
illocutionary devices. It goes without saying that languages differ from
one another in their 'illocutionary grammars'; but they also differ from
one another in the relative importance they give to this or that universal
device, and in the relative frequency with which this or that universal
device is used in a given language.
For example, it seems likely that in all languages which have the
imperative as a grammatical category, imperatives can be repeated 'for
emphasis'. Thus, one can say in English: Come in, come in! Run, rabbit,
run! Similarly, one can say in Japanese: Kaere, kaere! 'Go away, go
Discourse and illocutionary grammar 259
away!' In Italian, one might say: Parla, parla! 'Speak, speak!' Scappa,
scappa! 'Run away, run away!'
I think that repetition of this sort, which I shall henceforth call 'clausal
repetition', is different in function from the repetition of the entire
speech act, usually signalled in writing by the repetition of the same
'final' punctuation mark (a full stop, an exclamation mark, or a question
mark) rather than by the use of a comma; for example, sequences
such as: All right. All right., Mary! Mary! seem quite different in
function from their counterparts which would be transcribed with a
comma: All right, all right., Mary, Mary! What I want to suggest in
the present context is that in Italian, the use of 'clausal repetition' seems
to have a much wider scope, and a much greater importance, than it
has in English.
To give some substance to this claim, let me point out that in the
English translations of Italian novels numerous instances of 'clausal
repetition' (of imperatives, as well as other types of constituents) are left
out, or modified so that the overall level of direct repetition is reduced.
Consider, for example, the following intances of such a change:
Be.ne, bene, parleremo. (Manzoni 1972:131)
'Very well, we'll have our talk.' (Manzoni 1968:59)
Parla, parla ... (Manzoni 1972:160)
'Go on, speak out ...' (Manzoni 1968:75)
Vedra, vedra ... (Manzoni 1972:287)
'He'll see - he'll just see ... ' (Manzoni 1968:148)
Era indietro, indietro. (Manzoni 1972:134)
'Behind-hand, very much behind-hand ... ' (Manzoni 1968:60)
Ma senta, ma senta (Manzoni 1972:90)
'But listen, do listen ' (Manzoni 1968:36)
Ma ascolti, ma ascolti, ma ascolti. (Manzoni 1972:134)
'Listen, listen.' (Manzoni 1968:61)
Whatever the exact meaning of the 'clausal repetition' is, it would
appear that both the speakers of English and the speakers of Italian feel
on occasion a need to convey it (assuming that the meaning conveyed is
in both cases similar), but it also seems clear that the speakers of Italian
feel this need more often, and that the meaning in question plays a more
important role in Italian discourse than it does in English discourse.
260 Italian reduplication: its meaning and its cultural significance
What I would like to suggest is that there may be a link between the
enormous role of 'clausal repetition' in Italian discourse and the exis-
tence of the illocutionary device of 'syntactic reduplication' in Italian
grammar. It seems likely that the pragmatic meanings associated with
'clausal repetition' have led, through wide use, to the emergence of
a new grammatical category, a language-specific grammatical device
('syntactic reduplication'). In support of this suggestion, I would adduce
the fact that in many cases the same words seem to be among those
most frequently used in both repetition and reduplication:
Presto, presto! Presto presto!
'Quickly, quickly!'
Adagio, adagio! Adagio adagio!
'Slowly, slowly!'
3. The illocutionary force of clausal repetition
In English, the repetition of an imperative, or of an adverb which can
be interpreted as modifying an imperative, tends to be interpreted as
a directive urging the addressee to act immediately. Natural-sounding
utterances such as
Come in, come in!
Stop it, stop it!
Wait, wait!
Look, look!
Quickly, quickly!
all seem to imply a message which can be spelt out as 'I want you to
do something NOW'. Long-term directives sound less natural when
repeated:
?Look after yourself, look after yourself!
?Write to us, write to us.
One could say, of course:
Do write to us!
You must write to us!
Don't forget to write to us!
The illocutionary force of clausal repetition 261
but the repetition:
Write, write!
inevitably introduces a note of urgency and the implication: 'do it
right away'.
Furthermore, expressions which are compatible with a message of
urgency without necessarily implying it tend to acquire such implications
when repeated. For example, the expression all right can be used to
express simple agreement or acceptance; but all right, all right sounds
as if the speaker was trying to cut short the interlocutor's speech ('stop
saying it; I want you to do it now; I have already agreed, so there is no
need for you to go on talking about it').
In Italian, clausal repetition does not imply urgency, and so it is not
restricted to contexts where a message of urgency is appropriate. It
conveys a desire to influence the addressee ('I want you to do X') but
not necessarily to urge the addressee to do something quickly ('I want
you to do X now'). For example, these sentences:
In prigione, in prigione! (Manzoni 1972:347)
In una grotta, in una grotta; lontano da costoro. (Manzoni
1972:353)
convey long term goals, not urgent commands; it is appropriate, there-
fore, that in the English translations of Manzoni' s novel they have
been rendered in a modified form, without the perfect symmetry of the
Italian original:
To prison, yes, to prison! (Manzoni 1968:184)
A cave, a cave for me! Far from all this rabble. (Manzoni
1968: 188)
A fully symmetrical repetition would introduce a note of urgency,
absent from the Italian original.
But 'clausal repetition' can also be used - in both Italian and
English - in contexts where no goals (other than illocutionary ones)
are indicated, as for example in the following dialogue:
"Ben arrivato, ben arrivato!"
"Ben trovati."
"Avete fatto buon viaggio?"
"Bonissimo; e voi altri, come state?"
"Bene, bene."
(Manzoni 1972:407)
262 Italian reduplication: its meaning and its cultural significance
'Welcome, welcome!'
'Well met.'
'Have you had a good journey?'
'Excellent. And how are you all?'
'Fine, fine.'
(Manzoni 1968:223)
How do utterances such as 'welcome, welcome' or 'fine, fine' differ
in their communicative import from their hypothetical non-iterative
counterparts 'welcome' and 'fine'?
1 think one important difference lies in the degree of the speaker's
commitment to the utterance. Hearing 'X!' one can always remain
unsure whether the speaker really meant what he said, but hearing 'X,
X!' one should be able to rest assured that what was said was really
meant (or so the speaker implies). For example, when one says Thank
you one expresses gratitude, but it is possible to say it coldly, in a way
which will signal to the addressee that quite possibly no gratitude was
really felt. But if one says Thank you, thank you, one signals to the
addressee that one does mean what one says. For this reason, one cannot
say it coldly. (One can say it impatiently, dismissing the interlocutor,
but not coldly.) Of course when one says it warmly, one can do so
insincerely, pretending that one is grateful when in fact one is not, but
a cold or hostile tone is incompatible with the semantic component
encoded in the repetition as such.
Thus, whatever the illocutionary purpose of the first occurrence of the
repeated item might be (such as, 'I say it because 1 want you to know
it'), the illocutionary purpose of the second occurrence is distinct and
perhaps always the same:
1 say it another time because 1 want you to know that 'I mean it'
Accordingly, the following semantic representation can be proposed:
Come in, come in! Look, look! All right, all right!
(a) 1 say: I want X to happen
(b) I want it to happen now
(c) 1 know: you can think: 1 say this, 1 don't think this
(d) 1 want you to know: 1 think this
(e) I say this another time because of this
(f) I feel something when I say it
The illocutionary force of Italian reduplication 263
Component (b), which expresses something like urgency ('I want it
to happen now') is particularly easy to detect in imperative utterances,
such as Look, look! or Wait, wait!, where it refers to specific actions of
the addressee. It may be somewhat harder to detect in non-imperative
utterances, such as All right, all right! or No, no!, in which the speaker
may be seeking to influence the addressee's state of mind rather than
to prompt the addressee to any external action. But the impatience or
eagerness of such utterances shows that a reference to time ('now') is
present in them, too.
The assurance of something like sincerity ('you may think that I say
this, [but that] I don't think this; I want you to know that I think this') is
easy to detect in utterances which express good feelings towards the
addressee, for example in Welcome, welcome! or Thank you, thank you!,
and also in some imperative utterances, such as, for example, Come in,
come in! But can one really talk of 'sincerity' in the case of utterances
such as Look, look! or Wait, wait! which do not seem to similarly engage
the speaker's feelings? Probably not. Nonetheless, in urgent imperative
utterances of this kind it is also very important for the speaker to get
across the message: 'I mean it'; this can be interpreted, depending on the
context, either as 'I feel something (good towards you) - I mean it' or
as 'I want you to do it now - I mean it' (i.e. I mean now, not after now).
The proposed explication seems to fit all of these cases.
4. The illocutionary force of Italian reduplication
What is the difference in communicative import between instances of
clausal repetition, such as adagio, adagio 'slowly, slowly' and instances
of syntactic reduplication such as adagio adagio? And what is the differ-
ence between 'syntactic reduplication' and the 'intensification' marked
by words such as molto 'very'?
To begin with the first of these questions, I suggest that in both cases
(adagio, adagio and adagio adagio) the speaker insists on the truth or
validity of what is said. In the case of clausal repetition, the nature of
this insistence is unspecified: the speaker may be dismissing possible
doubts regarding the sincerity, seriousness, accuracy, or some other
aspect of the utterance. In the case of reduplication, however, the nature
of this insistence is quite specific: roughly speaking, it regards the
264 Italian reduplication: its meaning and its cultural significance
accuracy of the utterance. In calling someone' s eyes neri neri the
speaker insists that these eyes were 'really' black, literally black; that
their color was not close to black but exactly black; that no exaggeration
is involved. Saying subito subito the speaker insists that in saying 'at
once' what is meant is literally 'at once', and not something more or
less close to 'at once'; again, the reduplication implies that there is no
exaggeration in what is said. Saying 'barely barely' the speaker insists
that in saying 'barely' there is no exaggeration.
Furthermore, I conjecture that, unlike clausal repetition, reduplication
involves a single speech act. For this reason, I have assigned to redupli-
cation a single component in the frame 'I say', and a single illocutionary
purpose. In clausal repetition, the speaker performs a particular speech
act (such as a request) with its own illocutionary purpose (such as, 'I say
it because I want you to do it'), and then repeats the utterance, with a
new illocutionary purpose ('I say it another time because 1 want you
to know that I mean it'). In reduplication, no similar split seems to
occur. The prosodic unity of the reduplicated utterance mirrors, it seems
to me, its illocutionary unity, and this unity is reflected in the proposed
semantic representation.
Turning now to a comparison between the reduplication and the inten-
sification signaled by the word molto 'very', I would first of all reiterate
the observation that molto - unlike reduplication - is restricted to
gradable qualities. In essence, to say that someone is molto bella 'very
beautiful' or molto gentile 'very nice' is to say that the person in ques-
tion is more beautiful (more nice) than one could imagine on hearing
that she was simply beautiful (or nice). In the case of words which don't
refer to gradables the notion of 'more X' is simply not applicable. One
can't say *molto quasi 'very almost', just as one can't say *piu quasi
'more almost'.
What one can do in the case of non-gradables is stress that they are
being used responsibly, accurately, strictly. By repeating a word ('XX')
the speaker draws attention to that word, and insists on its strict
correspondence with what is meant ('I mean X, not something a little
different from X'). The fact that reduplicated expressions often co-
occur with the adverb proprio 'really, truly' (E proprio bianca bianca
'It is really white-white') is significant in this connection.
Of course one could still say that, when applied to gradable qualitative
adjectives and adverbs such as leggero 'light' or adagio 'slowly', the
reduplication signals a 'high degree' ('very light', 'very slowly'). But
there is no need to postulate a separate meaning for this use of redupli-
The illocutionary force of Italian reduplication 265
cation. Rather, we should say that the connotation of 'high degree' is
due to an implicature, calculable from the combined effect of the redupli-
cation and the invariant meaning of the base. For example, if adagio
adagio means something like 'the word adagio ("slowly") fits this
situation perfectly; in saying adagio I am not exaggerating', then it is
natural to infer that the process in question was in fact very slow. But
the general formula: 'no exaggeration' fits both qualitative and non-
qualitative uses of reduplication. There is no need, therefore, to analyse
this device as polysemous.
Furthermore, even in the case of gradable adjectives, reduplication
doesn't always produce an effect similar to that of molto 'very'. For
example (as Anna Ravano has pointed out to me), there was some years
ago a film entitled Un borghese piccolo piccolo 'A small-small citizen'.
The hero was a 'perfectly ordinary' man, who became involved in ex-
traordinary events. Clearly, the expression piccolo piccolo wasn't used
to mean 'very small', but to mean 'truly small, truly ordinary'. The stress
was not on a high degree of the feature in question, but on the accuracy
or validity of the word used.
The difference in function between reduplication (bella bella 'beauti-
ful beautiful') and 'intensification' (molto bella 'very beautiful') be-
comes even clearer when one considers instances in which reduplication
is applied to nouns rather than adjectives. For example, Lepschy -
Lepschy (1984:103) write: "Con i nomi l'intensificazione (0 meglio
un'identificazione della qualita autentica) si puo ottenere anche col
radoppiamento: caffe caffe, cioe caffe vero e non un surrogato." [With
respect to nouns, intensification (or, better, the identification of an
authentic quality) can also be achieved by means of 'doubling': caffe
caffe 'coffee coffee' is true coffee and not a surrogate.] Further examples
of nominal reduplication include brodo brodo 'broth broth', i.e. genuine
broth, and lana lana 'wool wool', i.e. genuine wool.
The formula offered by Lepschy and Lepschy ('the identification of
an authentic quality') is, I think, insightful. But the implication that the
device of 'doubling' (reduplication) has two different functions ('intensi-
fication' in the case of adjectives and 'identification of a true quality' in
the case of nouns) is unfortunate. In fact, the function of the reduplica-
tion is the same in both cases. But to capture it, we need a unitary
formula different from that used by Lepschy and Lepschy. Moreover, we
need a formula which would account not only for nouns such as caffe
caffe and for adjectives such as piccola piccola, but also for adverbs
and adver.bial expressions where no 'quality' is referred to, such as, for
266 Italian reduplication: its meaning and its cultural significance
example, subito subito 'at once at once' or quasi quasi 'almost almost'.
Before proceeding to tentative semantic representations, I would like
to draw attention to yet another difference between the two patterns
under discussion. I believe that expressions such as bella bella - unlike
expressions such as molto bella 'very beautiful' - contain an emotional
component, which can perhaps be represented roughly as 'I feel some-
thing thinking about it'.
For example, when one compares utterances like these:
a. Venga subito subito.
'Come at once at once.'
b
l
Come straight away.
b". Come at once - I mean at once!
one is bound to notice the highly expressive, emotional tone of (a), in
contrast to (b
l
) and (b"). It seems impossible to pronounce (a) with the
neutral, detached prosody with which (b') or (b") could be uttered. The
fact that the reduplication can be applied to purely descriptive adjectives,
such as nero 'black', bianca 'white', piccolo 'small', or fisso 'attentive'
doesn't undermine the hypothesis that the pattern contains an emotional
component. Examining utterances where expressions such as neri neri,
bianca bianca or duro duro are actually used, it is usually easy to detect
in the context clear clues to the emotional undertones. For example,
when one of the heroes of Manzoni' s novel quoted above undergoes a
great spiritual crisis and tosses and turns in his bed, unable to sleep, it
is small wonder that it seems to him that his bed has become duro duro
'hard hard' and his blankets, pesanti pesanti 'heavy heavy'.
Similarly, it is small wonder that in one of the most dramatic moments
of the story, an accidental witness, most intrigued, follows the amazing
dialogue fisso fisso 'attentively attentively'. In another dramatic mo-
ment, the hero, trying to escape from the police, wants to cross a river
in a fisherman's boat, without being noticed by anybody. Naturally,
therefore, he addresses the fisherman in a voice which is leggera leggera
'light light', 'soft soft'.
Examples can be multiplied. They all show, however, that the redupli-
cation adds an emotional dimension to the utterance. In some cases
(such as 'come at once at once') this aspect of meaning can easily be
captured by means of the component
I feel something thinking about it
(I feel something saying it?)
The illocutionary force of Italian reduplication 267
In others, the identity of the person who 'feels something' is less
clear: is it the hero who feels something, or is it the narrator, or both? I
am not sure. It seems to me, however, that if we wish to give a unitary
account of the use of the reduplication we should perhaps assume that
it is the speaker's feeling which is relevant to the pattern. In a story in
which the narrator merely empathises with the hero, the hero's emotions
may be primary and the narrator's secondary, but I am inclined to think
that it is the narrator's emotion which is signalled directly by the
reduplication.
1 suggest that the full illocutionary force of the reduplication can be
portrayed along the following lines:
[Her eyes were] neri neri 'black black'; [He was] ricco ricco
'rich rich'
(a) I say: X (her eyes were black; he was rich)
(b) 1 know: you can think:
I say 'X', I think: 'something like X'
(c) I want you to know:
I think 'X', not 'something like X'
(d) I say this like this because of this
(e) I feel something because of this
[I want you to come] subito subito 'at once at once'
(a) I say: I want X (I want you to come at once)
(b) 1 know: you can think:
1 say: 'X', 1 think 'something like X'
I want you to know:
I think 'X', not 'something like X'
(d) I say this like this because of this
(e) I feel something because of this
Instead of saying 'I say this like this because of this' in component (d)
(a phrasing suggested by Jean Harkins) one might consider an alternative
phrasing: 'I say this another time because of this'. It is not clear to me,
at this stage, which of these phrasings is more adequate.
As for the reduplicated nouns, such as caffe caffe 'genuine coffee', we
can assign to them virtually the same explication, without, however, the
emotional component (e):
caffe caffe 'coffee coffee'
(a) 1 say: X (coffee)
268 Italian reduplication: its meaning and its cultural significance
(b) I know: you can think:
I say 'X', I think 'something like X'
(c) I want you to know:
I think 'X', not 'something like X'
(d) I say this like this because of this
And for the 'intensification' marked by molto 'very', I suggested that
its pragmatic force could be spelt out along the following lines (see
Wierzbicka 1972:86):
Emolto ricco 'He is very rich'
(a) I say: he is X (rich)
(b) I want to say more than X
I have come to think, however, that 'very' may be a better candidate for
the status of a universal semantic primitive than 'more'. If this is correct,
then no further explication of words like molto or very would be neces-
sary. In any case, I don't posit for this pattern an emotional component
('I feel something thinking about it'), as I did for the syntactic redupli-
cation, because 'intensifiers' such as molto and very can be used in
contexts which exclude emotionally loaded words or expressions:
Objectively speaking, she is very beautiful.
?Objectively speaking, she is gorgeous.
?Objectively speaking, she is most attractive. (See 6,7 below.)
5. Clausal repetition as a means of 'intensification'
Bolinger (1972:90) has shown that repetition can be used in English as a
means of 'intensification', analogous to words such as very or extremely.
He adduced, among others, the following examples (which I think
should be qualified as gushing exaggerations):
That's very, very interesting.
They were quite, quite willing to accept.
She's a tiny, tiny baby.
It was a big, big bear.
He's a wonderful, wonderful person.
I carefully, carefully put it down.
Clausal repetition as a means of 'intensification' 269
As Bolinger points out, lexical iteration is accompanied here by
prosodic intensification (which Bolinger shows by spacing).
It seems to me that in sentences of this kind the repetition is, so to
speak, intraclausal, and that its function is quite different from what
I have called clausal repetition, exemplified by sentences such as:
Come in, come in.
Thank you, thank you.
Fine, fine.
In the case of clausal repetition, the speaker repeats a whole clause
(which may of course consist of just one word). In the case of intra-
clausal repetition, the speaker repeats a word. In the first case, the
speaker insists on the validity of what is said. In the second case, the
speaker insists that the WORD in question is well-chosen.
Intraclausal repetition is of course closer in function to syntactic
reduplication than is clausal repetition. Nonetheless, there are some
important differences between the two (intraclausal repetition and redu-
plication). For one thing, intraclausal repetition seems to introduce an
additional speech act, analogous, in a sense, to a parenthetical clause:
a. That's very, very interesting.
b. That's very interesting; I say: very (is the right word here).
a. I carefully, carefully put it down.
b. I carefully put it down; mark the word: carefully.
By contrast, Italian expressions such as neri neri seem to be used in
unitary speech acts (as expressions such as jet black are used in English).
Secondly, in English, intraclausal repetition seems to be confined
mostly to intensifiers (such as very and quite) and to words which have
an expressive component built into them. Purely descriptive words
don't lend themselves equally well to such use:
?She was a small, small baby.
?It was a large, large bear.
*?He's an intelligent, intelligent person.
The expression It's a small, small world (Jane Simpson, p.c.) sounds
better than a small, small baby, precisely because the former, in my
view, encodes an emotional component, and is usually uttered with a
wordly-wise smile or some other expression of feeling. Expressions
such as neri neri or in fretta in fretta indicate that there is no similar
restriction in Italian.
270 Italian reduplication: its meaning and its cultural significance
Third, English intraclausal repetition is restricted to contexts com-
patible with the intended meaning of 'intensification'. One can hardly
say in English
?/ almost, almost apologised to him.
?Make just, just enough room for us to be able to pass.
Above all, however, it must be stressed that syntactic reduplication is
an illocutionary device which Italian has grammaticalised; intraclausal
repetition, however, may well be a rhetorical device available (though
not necessarily widely utilised) in any language. The exact meaning, and
the exact range of use, of syntactic reduplication cannot be calculated
on the basis of any universal principles of human behavior (such as
Grice's maxims). It is part and parcel of Italian grammar, which doesn't
have counterparts in English, Japanese, or even French.
Of course it cannot be denied that the basis of this device is largely
iconic, and certainly reduplication is known to convey related mean-
ings in very many unrelated languages of the world (see, for example,
Moravcsik 1978). But reduplication can also be put to entirely different
uses (see, for example, Hayakawa 1985; Wilkins 1984; for a broad
survey, see Mel'cuk, in press). The exact force of Italian reduplication
is at least partly a matter of language-specific conventions, and so it
cannot be regarded as due to 'implicature' rather than meaning.
6. The absolute superlative in Italian and in English
Before turning to the cultural significance of 'syntactic reduplication',
I think it in order to discuss briefly another grammatical device which
is also characteristic of Italian and whose pragmatic force is closely
related to that of 'raddoppiamento'. Italian grammarians call this device
superlativo assoluto, the 'absolute superlative'. It can be illustrated with
forms such as devotissimo 'extremely devoted', bianchissimo 'extremely
white', or velocissimo 'extremely fast'.
The superlativo assoluto continues, historically, the old Latin superla-
tive. But from a synchronic point of view, it is a separate grammatical
category, distinct from the (Italian) superlative; that is, from what Italian
grammarians call superlativo relativo:
The absolute superlative in Italian and in English 271
comparativo:
superlativo relativo:
superlativo assoluto:
bello
piu bello
il piu bello, bellissimo
As one grammarian (Fochi 1966: 168) puts it: "quando, per esempio,
diciamo che 'Giulio e generosissimo' (superlativo assoluto) non con-
sideriamo quanto siano generosi, al suo confronto, Tizio, Caio e via
dicendo: osserviamo tale qualita in lui solo, e ci basta affermare che
egli la possiede in grado molto alto." [When we say, for example, that
Giulio is 'most generous', we are not comparing Giulio' s generosity
with that of Tizio, Caio, and so on: we are considering the quality ex-
clusively in him, and all we want is to affirm that he possesses it in
a very high degree.]
Italian grammars commonly describe the 'absolute superlative' as a
device equivalent to reduplication, and some even extend the same label
(superlativo assoluto) to both of these devices (see, for example, Fochi
1966:168; Kaczynski 1964:106). In fact, however, there are important
differences between the two devices (as well as similarities). For one
thing, the absolute superlative in the strict sense of the term is, like molto
'very', restricted to qualities, and to qualities regarded as gradable. One
cannot say, for example, subitissimo as one says subito subito 'at once
at once'. For another, the absolute superlative is not meant to convey
accuracy. Normally, it involves a self-evident exaggeration; and this
exaggeration is functional, in view of the speaker's emotional attitude.
For example, if one describes a drink as una bevanda agrissima 'most
bitter' (Fochi 1966: 170) or if one describes an apple as una mela
asprissima 'most sour', one is not pretending to be accurate: the very
exaggeration serves to highlight the speaker's displeasure.
On the other hand, the reduplication does lay a claim to precision; and
for this very reason, it is inappropriate in purely emotional contexts,
where no descriptive content is conveyed. For example, the form caris-
simo 'dearest' is used in Italian very frequently; but the hypothetical
form carD carD sounds comical. A repetition with a comma intonation:
caro, carD sounds all right, of course (cf. in English I want you to meet
a dear, dear friend of mine) but a reduplicated carD carD sounds odd.
Similarly, forms such as illustrissimo 'most illustrious', or obbligatis-
simo 'most obliged' are commonplace; but forms such as illustre illustre
or obbligato obbligato sound peculiar. This peculiar effect has nothing
to do with the sincerity, or otherwise, of the alleged emotion. When the
emotion expressed is purely a matter of social convention there is still
272 Italian reduplication: its meaning and its cultural significance
a clash between the purely subjective, non-descriptive content of the
adjective and the purported stress on precision conveyed by the redupli-
cation; as a result, forms such as obbligato obbligato or illustre illustre
sound as odd as caro caro.
To capture the relevant semantic relationships, I would propose a
semantic representation along the following lines:
[She is] gentillissima 'extremely nice'; [It is] velocissimo
'extremely fast'
(a) I say: it is very X
(b) no one/nothing could be more X
(c) I feel something thinking about it
The comparative component (b) could be further explicated as follows
(cf. Wierzbicka 1971):
no one/nothing could be more X =
if one can't say this of this person (thing)
one can't say this of anybody (anything)
One could say that English, too, possesses a category of an 'absolute
superlative', and that it can be illustrated with expressions such as most
kind, most generous, most helpful, most unpleasant, and so on. It must
be pointed out, however, that in English the use of this category is
very restricted. For example, we may be able to call someone 'most
kind', but we couldn't call a thing 'most white' or 'most long' (in the
absolute superlative sense). But in Italian, the forms such as lunghissimo
or bianchissimo are perfectly acceptable.
One generalisation which suggests itself is that in English, the pattern
'most Adj.' cannot be applied to purely descriptive adjectives, that it
applies only to terms of assessment ('good' or 'bad'). However, to say
this is not sufficient, as the following contrasts in acceptability show:
He was most helpful.
It was most kind of you.
She is most attractive.
I am most grateful.
She has a most attractive
personality.
It was most effective.
It was a most generous offer.
It was most disappointing.
?He is most good-looking.
?She is most beautiful.
??She is most healthy.
?He was most pleased.
??She has most regular
features.
?It was most elegant.
??It was a most silly play.
??It was most bad.
The absolute superlative in Italian and in English 273
It was most frightening.
It was most ingenious.
??It was most deafening.
??He is most modest.
Studying contrasts of this kind, I am led to believe that the pattern
'most Adj.' refers to an effect that people, and human acts, have on other
people. In fact, even combinations of the word most with 'human pro-
pensity' adjectives such as kind, generous, or helpful (see Dixon 1977)
sound most natural (in the relative, not the absolute, superlative sense!)
in sentences referring to specific acts of kindness or generosity, rather
than in abstract 'character references'. In the two columns below, sen-
tences on the left sound, I think, more felicitous than those on the right:
It was most kind of you.
He was most generous (to us,
in his dealings with us).
He was most unpleasant (to us).
John is most kind.
Mary is most generous.
Max is most unpleasant.
If the adjective in the absolute superlative construction refers to a
person rather than to an act, it sounds best in the construction which
Bolinger (1977:141-142) calls 'the ergative of (it was Adj. of you/him),
which implies an evaluation of a person in relation to some specific act
performed by him or her.
However, expressions such as most attractive, which don't imply
actions, demonstrate that interaction between people is not a sine qua
non of this construction; an effect of a person on other people is suffi-
cient for the pattern to be able to be used. Near-synonyms such as attrac-
tive and beautiful or good-looking are particularly instructive in this
respect: a person's good looks may leave other people cold; but a
person's attractiveness, by definition, cannot. Similarly, it is significant
that most grateful sounds much better than most pleased - presumably,
because the word grateful refers to human interaction and implies an
emotional reaction to another person, whereas the word pleased has no
such implications.
However, one cannot neatly divide the set of English adjectives
into those which always can, and those which never can, occur in the
absolute superlative construction. The real constraint is semantic, not
lexical: it is the meaning of the entire sentence, not just of the adjective
as such, which determines the degree of the sentence's acceptability.
For example, of the two sentences below:
a. She is most beautiful.
b. He gave me a most beautiful necklace.
274 Italian reduplication: its meaning and its cultural significance
(b) is more acceptable than (a), presumably because it refers to human
interaction and implies an emotional reaction to it (gratitude).
Quirk et al. (1972:287) assert that a sentence such as She is most
beautiful "can only mean she is extremely beautiful and not that she
is more beautiful than all the others". But if an expression such as
most beautiful really didn't mean more than the expression extremely
beautiful then we would have no explanation for the gushing, exagger-
ated character of the former, in contrast to the latter. I think that
Jespersen's observation that expressions such as most beautiful are
exaggerated should not be lost sight of. And we do lose sight of it if we
say that most means no more than extremely.
It is not an accident that the absolute superlative shares, to some
extent, the morphology of the relative (comparative) superlative. A com-
parison is there, in both cases. The differences lie elsewhere: in the
nature of the comparison, and in the illocutionary purpose of the utter-
ance. If I say: Mary is the most attractive girl in her class, I am com-
paring Mary to a specific, finite set; and I am informing the addressee
that I set her above all the other members of that set. By contrast, if I
say: Mary is most attractive, I am comparing Mary to an imaginary set
of competitors; and if I am setting her above those imaginary competi-
tors it is not in order to inform the addressee ('I say it because I want
you to know it') but in order to express my emotional reaction to my
perception of Mary ('I feel something because of it').
Quirk et al. (1972:287) also say that "absolute most is restricted as to
the adjectives with which it occurs, perhaps premodifying only those
expressing subjective rather than objective attitudes ... : She is most
unhappy; *She is most tall." I think that the basic point is correct, but
that the way the restriction is formulated is clearly not sufficient. I pre-
sume that what the authors really have in mind is not so much a contrast
between 'subjective attitudes' and 'objective attitudes' as a contrast
between attitudes and objective characteristics (clearly, being tall is not
an 'objective attitude'). But the statement they offer does not explain
why there is a difference in acceptability between, say, unhappy and sad,
or unhappy and pleased, or unhappy and happy:
She is most unhappy.
?She is most angry.
*She is most happy.
I think the clue to these differences in acceptability lies partly in the
emotional effect that a person's 'subjective attitude' exerts on other
The absolute superlative in Italian and in English 275
people. Seeing a person who is obviously very angry we may fail to 'feel
something because of that'; and another person's extreme happiness is,
unfortunately, unlikely to echo in the hearts of the bystanders; but seeing
someone who is obviously very unhappy, we are less likely to remain
totally indifferent.
It seems to me that the reason why the range of use of the English
'absolute superlative' is so much more restricted than that of its Italian
counterpart is that the meaning encoded in the English construction is
much more specific than that encoded in its Italian counterpart. X-issimo
seems to imply, 'it is more X than anything or anyone that one could
imagine', but X-est or most X seem to imply more: 'it is more X than
anyone or anything that one could imagine', and also 'this is good/bad
for someone'. One might say, then, that the Italian absolute superlative
constitutes a grammatical device which enables the speakers of Italian
to perform a kind of expressive overstatement any time, that is to say
regardless of the nature of the qualities spoken of, whereas English is
much more restrictive in this respect.
Another, related, difference concerns the nature of the emotional
component embodied by both the Italian and the English construction. In
English, what is involved is, typically, an emotional reaction to a human
act, an act which can be evaluated. (Another English quasi-superlative
construction, which doesn't take the definite article, seems to be re-
stricted to the expression of emotions: My deepest sympathy, With best
wishes, With warmest regards.) If it is not an act, then it is a perception,
followed by an assessment and combined with an emotional reaction.
In Italian, the nature of the emotional component is not similarly
restricted. Forms such as velocissimo 'extremely fast' or nuovissimo
'extremely new', 'completely new' can be used in abstract descriptions
of objects, where there is no question of a direct reaction to an act, or
even to a perception. The difference in question is subtle, but I think we
should at least try to capture it in the semantic representations of the
two constructions, to account as accurately as possible for the differ-
ences in the range of their use.
I would propose the following way of portraying this difference:
You are most generous; I am most grateful; She is most
attractive =>
I feel something because of this (i.e. because of what you have
done, because of the way she looks, etc.)
276 Italian reduplication: its meaning and its cultural significance
E nuovissimo (velocissimo, bianchissimo, etc.) =>
I feel something thinking about it (i.e. thinking of how new it
is, etc.)
The two full explications would then compare as follows:
E velocissimo/bianchissimo (It is 'most fast', 'most white')
(a) I say: it is very X
(b) nothing could be more X
(i.e. if one cannot say 'X' of this one cannot say it of
anything)
(c) I feel something thinking about it
It is most ingenious/unpleasant.
(a) I say: it is very X
(b) nothing could be more X
(i.e. if one cannot say 'X' of this one cannot say it of
anything)
(c) I think: this is good/bad for someone
(d) I feel something because of this
7. Illocutionary grammar and cultural style
Different cultures favour different styles of social interaction, and illocu-
tionary grammars tend to reflect cultural differences of this kind. For
example, the fact that in Anglo-Saxon culture respect for a person's
autonomy occupies a high position in the hierarchy of values is reflected
in the great importance given in this culture to pragmatic values such
as 'tact', 'non-interference' and 'anti-dogmaticism'; this, in turn, is re-
flected in the exuberant growth of interrogative and quasi-interrogative
devices in the grammar of English (see Goody 1978; Chapter 2 above).
The importance of the illocutionary strategy of understatement in
English speech (see Hubler 1983) also has its obvious roots in Anglo-
Saxon culture. The fact that in English, understatement can be used even
in those situations where the speaker wishes to speak in the strongest
possible terms, is highly significant in this respect. For example, one can
say in English that a crime was 'rather horrible', or that a performance
was 'pretty awful', or that a student is 'fairly enthusiastic' - all expres-
Illocutionary grammar and cultural style 277
sions which cannot be literally translated into languages such as Polish
or Italian, whose speakers prefer emphatic overstatement to cautious
understatement (see Chapter 2). For example, expressions such as abbas-
tanza orribile 'rather horrible' or piuttosto orrendo 'pretty awful' sound
rather ludicrous in Italian.
Of course one cannot identify Anglo-Saxon culture with the English
language. There are many cultural divisions within the English-speaking
world, and English is spoken today as a first language by many
groups who do not belong to the Anglo-Saxon cultural tradition. But
a language reflects past traditions, as well as the living culture. Since
the Anglo-Saxon tradition is, or has been for a long time, the dominant
one in English-speaking communities, it is little wonder that this tradi-
tion has left a strong imprint on the English language. But it should also
be pointed out that sooner or later, historical and cultural diversity within
the English-speaking world does find its reflection in linguistic diversifi-
cation. For example, the English spoken by American Blacks differs
from the mainstream in ways which are culturally revealing (see, for
example, Abrahams 1970, 1974; Kochman 1972; Mitchell-Kernan 1971,
1972; see also Chapter 3 above). And Australian English has developed
many distinctive features which reflect Australian history, culture, and
ethos (see Wierzbicka 1986b; Harkins 1988; Chapter 5 above).
I would like to suggest that syntactic reduplication and the absolute
superlative belong to a system of illocutionary devices which reflect,
jointly, certain characteristic features of Italian culture and, in particular,
of the Italian styIe of social interaction.
In my view, the absolute superlative constitutes, in a sense, an antithe-
sis of understatement. Discussing expressions such as most kind and
most ingenious, Jespersen (1965, 7:395) attributes them to a 'universal
tendency to exaggerate'. Leech (1983: 147), too, regards hyperbole as a
'natural tendency of human speech'. However, different cultures differ in
the extent to which they encourage, or discourage, this 'natural human
tendency'. The contrast between the English and Italian absolute superla-
tive constructions, provides, I think, a good illustration of the more
general differences between the two cultural traditions and cultural
styles. The device of syntactic reduplication, too, can be regarded as the
antithesis of understatement, and as a case of emotional overstatement.
But concepts such as 'understatement', 'hyperbole', or 'litotes' have
been taken over directly from traditional rhetoric, and, although they are
useful as hints and points of orientation, they are nonetheless vague
and imprecise, as most such traditional notions have had to be.
278 Italian reduplication: its meaning and its cultural significance
The use of explicit formulae, which force us to be rigorous and
explicit, helps us to recognise that expressions such as neri neri or duro
duro don't necessarily involve 'exaggeration' in the sense in which
expressions such as most ingenious or bellissima do.
The absolute superlative can be plausibly interpreted as meaning 'it
is more X than one could imagine'. The speaker is fully aware of saying
'a little more than what is strictly true', and is not trying to deceive the
addressee on that score. The speaker assumes that it will be clear to the
addressee that 'a little more than what is true' is said in order to convey
an emotional attitude to the state of affairs spoken of.
In the case of syntactic reduplication, the speaker is doing something
different. This time, he does insist that what he says is not even a little
different from what is true. And yet this strategy, so different from that
of emotional exaggeration, nonetheless has a similar pragmatic effect
(that of 'overstatement', or 'anti-understatement'). I think that in order to
understand this apparent paradox, it is useful to compare expressions
such as bella bella not only with expressions such as bellissima but
also with English expressions such as 'rather pretty'. The person who
says (see also Leech's 1983:148 examples):
She's rather pretty.
We're rather proud of it.
Actually, I'm rather good at it.
is trying hard not to say 'very pretty', 'very proud', or 'very good' and
clearly conveys this message to the addressee ('I don't want to say more
than 'pretty', 'proud', 'good'.') The addressee can infer that the speaker
thinks more than he says ('he says that he doesn't want to say more than
X - presumably, he thinks more than X'). In this sense, the English
strategy in question can indeed be interpreted as an understatement. It
is worth noting also that utterances of this kind are often introduced
with an apology: I don't want to boast, but ....
The speaker who says bellissima or velocissimo does say more than
what the bare adjective would convey, and clearly conveys his intention
of doing so ('I want to say more than X'). This, however, is not different
from the effect conveyed by 'intensifiers' such as molto or very. The real
'overstatement' comes, I think, in the implicit comparison: 'more than
'one could imagine'.
The person who says bella bella doesn't imply 'more than beautiful',
let alone 'more beautiful than one could imagine'. But his behaviour,
too, is (though in a different sense) the opposite of that of somebody who
Illocutionary grammar and cultural style 279
says 'rather pretty'. The person who says bella bella insists on the
absolute validity of his words; the person who says 'rather pretty'
deliberately refrains from doing so. The first one emphatically dots the i,
confident of the absolute validity of what he says and ready to assert
himself to the full, without 'tactfully' anticipating the possibility of
different views. The person who says 'rather pretty' doesn't want to dot
the i, doesn't want to insist on the absolute validity of what he says,
and does wish to leave room for other points of view.
Of course Anglo-Saxons, too, can and do 'overstate', when they wish
to do so - both in the sense of saying more than they really mean, and
in the sense of insisting on the absolute validity, and the absolute accu-
racy, of what they are saying (see Jespersen 1965,7:395; Sapir 1949:145;
Leech 1983:145; Brown - Levinson 1978:224). But the fact that, unlike
Italian, English has no grammatical devices for doing so suggests the
existence of a cultural difference - presumably, the same cultural
difference which is also reflected in the different scopes of the English
and Italian absolute superlatives.
I would add that the emotional nature of expressions such as bellis-
sima or bella bella seems as significant, from a cultural point of view,
as their 'emphatic' and 'overstated' character. As I have argued in
Chapter 2, there is a link between the rich system of expressive deriva-
tion (diminutives, augmentatives, and the like) in languages such as the
South Romance or Slavic ones, and the uninhibited display of emotions
characteristic of Mediterranean and Slavic cultures. There is also, I be-
lieve, a link between the virtual absence of expressive derivation in
English and the taboo on overt displays of emotion in Anglo-Saxon
culture. In English, utterances like You're too, too kind, She gave me the
most beautiful ring, He's a dear, dear man (Jane Simpson, p.c.) are
associated with a speech style ascribed to rich women, private-school
girls, homosexuals, actors, etc., who are popularly supposed to engage in
public displays of emotion, especially affection, or hysteria. And these
are thought of as insincere.
The fact that both syntactic reduplication and the absolute superla-
tive, so characteristic of Italian, embody an emotional component, is, I
think, another manifestation of the same cultural difference.
But while an uninhibited display of emotions paralleled by a rich
system of expressive derivation is as characteristic of, say, Russian or
Polish as it is of Italian (in fact, more so), devices such as syntactic
reduplication don't have any counterpart in Slavic languages - certainly
not on the same scale. Russian and Polish are at least as 'emotional' as
280 Italian reduplication: its meaning and its cultural significance
Italian; and when compared with English, they must be characterised as
favouring 'overstatement' rather than 'understatement' (at least in the
expression of opinions). Nonetheless, they don't have, or don't make
much use of, pragmatic devices such as syntactic reduplication.
As for the absolute superlative, Polish doesn't have it, whereas Rus-
sian does (see Mel'cuk, in press). This difference can be seen as related
to the fact that derivation of expressive diminutives goes much further in
Russian than it does in Polish (for example in Russian, but not in Polish,
diminutive suffixes can be added to relative adjectives and adverbs, ~ u
as pervyj 'first', pravyj 'right', or casto 'often'; the same holds for many
categories of nouns). Furthermore Russian, in contrast to Polish, has at
least some expressions analogous to reduplicated expressions in Italian,
such as cut' -cut' 'a bit' (lit. 'a bit a bit'), net-net 'rarely and unpredicta-
bly' (lit. 'no-no'), vot-vot 'is on the point of' (lit. 'there-there'). Genuine
syntactic reduplication is very rarely used in Russian, but it is not
impossible, as the following examples demonstrate:
No vy ne mozete ze menja seitat' za devocku, za malen' kuju-
malen' kuju devocku, posle moego pis' ma s takoju glupoju sutkoj!
(Dostoevskij 1976:167)
'But you can't consider me as a child, a little girl, after that
silly joke!' (Dostoevsky 1974:186)
fa dumala on takoj ueenyj, akademik, a on vdrug tak gorjaco-
gorjaco ... (Dostoevskij 1976: 178)
'I thought he was so learned, such a savant, and all of a sudden
he behaved so warmly ... ' (Dostoevsky 1974:198)
Ona vidimo cego-to stydilas' i, kak vsegda pri etom byvaet, by-
stro-bystro zagovorila sovsem 0 postoronnem. (Dostoevskij
1976:195)
'She was evidently ashamed of something, and, as people always
do in such cases, she began immediately talking of other things.'
(Dostoevsky 1974:217).
Can one identify any specifically Italian cultural features which might
explain the role of syntactic reduplication in Italian speech? Looking
for such features, we might recall the 'theatrical quality' of Italian
life (Barzini 1964:73), the "importance of spectacle", "the extraordinary
animation ... , the expressive faces, the revealing gesticulation", and "the
noise", which are "among everybody's first superficial impressions in
Italy, anywhere in Italy, in the north as well as in the south, in big cities
Illocutionary grammar and cultural style 281
as well as in decrepit and miserable hamlets forgotten by history"
(Barzini 1964:66).
The point about 'noise' is perhaps particularly relevant in the present
context. Barzini writes:
The noise is usually deafening. People chat, whistle, swear, sing, curse,
cry, howl, weep, call to each other and shout, carrying on elaborate discus-
sions or delicate negotiations. Mothers murmur endearing baby words to
their little children and ask bystanders to be witnesses to their darlings'
charm and pigheadedness. Other mothers call their sons from top-storey
windows with voices carrying to the next province. Bells clang with deep
bronze notes from the top of the belfry above, drowning every other sound.
Somebody is always practising the cornet or the trombone. At times the
same popular song or famous operatic aria comes apparently from every-
where, from radios in every shop, from the open windows of apartments,
from under the tables in the cafe, from the pockets of clients, from the
abdomens of passing housewives. Vespas, cars, motorcycles, trucks go by
with roaring engines.
The air is in fact filled with so much noise that one must usually talk in
a very loud voice to be understood, thereby increasing the total uproar.
Lovers sometimes have to whisper 'I love you' to each other in the tones
of newsboys selling the afternoon papers. Italians on their death beds, in
rooms facing especially noisy squares or streets, are known to have re-
nounced leaving their last wishes and advice to weeping relatives, being
too weak to make themselves heard. It is, however, a gay and happy noise,
magnified by the stone walls, the absence of greenery, the narrowness of
the streets. It goes on from dawn to the small hours of the night, when the
last strollers stop under your bedroom window to debate a fine point of
politics or the personality of a common friend, both speaking at the same
time at the top of their voices. (Barzini 1964:59-60)
What applies to 'noise' applies also to facial expressions and to
gestures:
What makes all such scenes more intensely fascinating is perhaps the
transparency of Italian faces. Conversations can be followed at a distance
by merely watching the changing expressions of those taking part in them.
You can read joy, sorrow, hope, anger, relief, boredom, despair, love, and
disappointment as easily as large-printed words on a wall poster. ...
Then there are the gestures. Italian gestures are justly famous. Indeed
Italians use them more abundantly, efficiently, and imaginatively than
other people. They employ them to emphasise or clarify whatever is said,
to suggest words and meanings it is not prudent to express with words,
sometimes simply to convey a message at great distance, where the voice
could not carry. . ..
282 Italian reduplication: its meaning and its cultural significance
A few gestures are as arbitrary and conventional as the deaf and dumb
alphabet or the sign language of American Indians. Most of them, how-
ever, are based on natural and instinctive movements, common to the
majority of men, certainly common to all Western men, elaborated,
intensified, stylised, sharpened, made into art. Like all great traditional
arts, this one too can generally be understood by the inexperienced at first
sight. (Barzini 1964:61-62)
It seems to me that this 'loudness', this animation, this display of
Italian life go a long way towards explaining the relevance of a
pragmatic device such as syntactic reduplication to Italian culture. Just
as Italian gestures, facial expressions, and vocal productions, can be
seen as 'exaggerated', hyperbolical, emphatic, and dramatic (as well
as emotional), so can Italian pragmatic devices. Sounds, gestures,
and facial expressions are 'overdone' - for purposes of self-expres-
sion, of theatrical display, and of dramatic effect. (It is also worth
recalling, in this connection, the pragmatic principle of 'interest',
posited by Leech 1983:146 in connection with overstatement.) Linguistic
devices such as the 'absolute superlative' and the syntactic reduplica-
tion can be seen as part and parcel of this characteristically Italian
cultural styIe.
8. Conclusion
It is not my present purpose to discuss in any detail the use of the
'absolute superlative', or any other illocutionary devices related to
syntactic reduplication. Rather, I want to reiterate the general point,
that illocutionary devices characteristic of a language are not mutually
independent, but tend to form networks of 'conspiracies' aimed at
common cultural targets.
At one time, scholars referred to such conspiracies using old-fash-
ioned labels such as Sprachgeist, the 'spirit of language' (see, for
example, Humboldt 1903; Vossler 1904, 1925; Spitzer 1928). When a
sociological and anthropological orientation replaced the earlier philo-
sophical and psychological bias, linguists started to talk about 'language
as a guide to social reality' (Sapir 1921; Whorf 1956) rather than about
'language as an expression of Volksgeist'. Later, the notion of 'cognitive
style' became an acceptable way of referring to more or less the same
Conclusion 283
sorts of phenomena (see Hymes 1961). These days, the preferred con-
ceptual umbrella is that of 'cross-cultural pragmatics' (for example,
Pride 1985).
I am not suggesting that these changes in ways of speaking are purely
superficial, and that they don't reflect any deeper changes in concerns,
assumptions, and methodologies. I believe, however, that it is important
to see the present concerns in the area of what is now called cross-
cultural pragmatics in their proper historical perspective: to recognise the
continuity of the tradition, and to learn how to learn from past insights
as well as from past mistakes; and, above all, to sharpen our methodo-
logical tools so that 'cross-cultural pragmatics' will come to represent a
real rather than a purely nominal progress with respect to the writings
of our predecessors who concerned themselves with similar problems
a century or half a century ago.
It seems to me that real methodological progress can be achieved by
translating the problems of cross-cultural pragmatics into the language
of illocutionary semantics. What our predecessors lacked was, above all,
methodological rigor and conceptual discipline. They lacked a rigorous
analytical framework which would allow them to study both the similari-
ties and the differences between the languages compared (and, for that
matter, between related constructions within one language) with a clear
sense of purpose and with clear standards of precision.
Today, in a post-structuralist and post-Chomskyan era, it is widely
felt that new standards of explicitness and rigour (if not of formalisation)
are called for, in this area of linguistics as in others. But rather than try
to develop and to sharpen new methodological tools which would
allow them to meet the required standards, many linguists prefer to
abandon the vital questions concerning links between language and
culture altogether. Certainly, by avoiding such questions they avoid
many errors and blunders which one might commit when one ventures
into this 'unsafe' territory. But while preventing themselves from
committing many errors and blunders they are also preventing them-
selves from discussing worthwhile questions and perhaps reaching
worthwhile insights. They are, in other words, narrowing the horizons
of linguistics and making it less exciting and less relevant to vital
human concerns.
Certainly, any discussion of the relation between illocutionary
grammar and cultural style should be carried out in an extremely careful
and cautious way. I suggest that the use of a semantic metalanguage
suitable for a standardised description of 'pragmatic' and illocutionary
284 Italian reduplication: its meaning and its cultural significance
meanings can provide a partial answer to the question of how to
combine rigour with insight in this tantalising area.
Chapter 8
Interjections across cultures
Pnin sighed in Russian: Och-och-och!
(Nabokov 1983:105)
1. Preliminary discussion
1.1. Interjections: phjsis and thesis ('nature' and 'convention')
Serge Karcevski (1969 [1941]: 196) starts his pioneering 'Introduction to
the study of interjections' with the following quote from a novel by
Alexandre Dumas:
Aha! - s' ecria-t-il en portugais.
'Aha! - he exclaimed in Portuguese.'
Karcevski notes that this sentence has an unintentional humorous effect,
and he attributes this effect to the fact that "one doesn't doubt for a
moment that we have to do here with 'natural' language, which doesn't
have to be learnt". But in fact, Karcevski points out, "to have the right
to think that, one should first engage in systematic investigations of the
functioning of interjections in all the remotest corners of the Earth".
The need for such systematic investigations is of course undeniable.
But even without such investigations we know - as Karcevski was
very well aware - that interjections differ considerably from language
to language. In fact, far from being universal and 'natural' signs which
don't have to be learnt, interjections are often among the most character-
istic peculiarities of individual cultures. Rosten (1968:26) makes this
point about the Yiddish interjection nu, which, he says, "is the word
most frequently used (aside from 'oy' [another interjection] and the
articles) in speaking Yiddish". According to Rosten, "Nu is so very
Yiddish an interjection that it has become the one word which can
identify a Jew. In fact, it is sometimes used just that way, i.e. instead of
286 Interjections across cultures
asking, 'Are you Jewish?', one can say, 'Nu?' (The answer is likely to
be 'Nu-nu. ')"
This suggests that interjections are based largely on culture-specific
conventions, not on any universal 'laws of nature', or at least that some
mixture of the two is usually involved.
In her charming Body book Claire Rayner writes:
Smells travel in the air, and go into your nose when you breathe.... Your
brain thinks about the smell. . .. If it is a bad smell your brain makes you
move your nose away from it. It tells you to wrinkle your nose, or cover
it with your hand. It sometimes makes you try to blow away the air
which has the bad smell in it. That's why people say 'Pooh!' to bad
smells. It is a blowing-away sort of word. (Rayner 1978:21)
This sounds quite convincing, but it doesn't change the fact that speakers
of languages other than English don't say Pooh! when confronted with
a bad smell, although they might say something similar. For example,
in German they say Pfui! and in Polish, Fu! (See below, section 3.)
Clearly, there are some similarities, but there are also differences. Gener-
ally speaking, interjections often show remarkable similarities across
language and culture boundaries, but these similarities are unpredictable
and have to be learnt just as much as the differences. Often, there are no
similarities at all. For example, the English interjections gee and wow
have no equivalents or near-equivalents in Polish (my own native lan-
guage), and a Polish immigrant in an English-speaking country has no
way of guessing what they mean and how they are to be used. On the
other hand, the English interjections ha or ow do have their counterparts
in Polish (similar in both form and meaning). But Polish immigrants
cannot predict that, and have to learn these interjections as much as
they have to learn gee and wow.
Usually, dictionaries recognise this unpredictable and conventional
character of interjections by including them (or some of them) and trying
to define them. But the definitions they offer are not of the kind that
could help anyone to learn how to use them. For example, if LDOTEL
(1984) says that in English, wow is "used to express strong feeling
(e.g. pleasure or surprise)" and that gee is "used to express surprise or
enthusiasm"; it gives the reader no clue whatsoever as to how these two
interjections differ in use from one another (and what is, for example, the
difference between a gee of surprise and a wow of surprise). Similarly,
when it says that ah is "used to express delight, relief, regret or
contempt" and that phew is "used to express relief, amazement or
Preliminary discussion 287
exhaustion" it gives no indication as to the difference between an ah
of relief and a phew of relief. Furthermore, no definitions of the kind
provided by LDOTEL (or by other conventional dictionaries) would
teach the reader in what situations wow, gee, ah or phew would be
inappropriate. And yet there is no way a language learner could guess
this, without instruction, or without prolonged immersion in an English
speaking environment.
But if interjections do not belong to any 'natural language' and have
to be learnt, why is it that they sound comical if reported in direct
discourse?
The first observation which should be made at this point is that direct
discourse in general is somewhat difficult to report in the frame 'X said
in language L'. It seems a little odd to say (or to write) the following:
??Good morning - she said in Russian.
??How nice to see you - she said in French.
?It is getting late - she whispered to him in German.
?Who is this? - he asked her in Italian.
The reason for this slightly odd effect is that (as I have suggested in
Wierzbicka 1974) direct discourse represents something like play-acting;
the addressee is invited to imagine that the 'reporter' imitates the
original speaker's utterance, and therefore that he reports jointly the
what and the how. On the other hand, the phrase 'in language L' indi-
cates that the reported speech separates the what from the how, and that
what is being reported is the content but not the manner.
Nonetheless it seems clear that some types of utterances are more
felicitous in the kind of reported speech under discussion than others. In
particular, exclamations involving complex expressions such as Good
heavens! are somewhat more felicitous in this context than primary
(global) interjections such as Oops!, Ouch! or Wow!; lexically produc-
tive exclamations such as How nice! or what a woman! sound a little
more felicitous than lexically restricted ones such as good heavens!;
and full questions or declarative sentences sound more felicitous than
exclamations of any kind. For example, informants tend to agree to the
following gradation in acceptability:
?It is time to go - she said in Russian.
??What a woman! - she said in Russian.
?? ?Good heavens! - she said in Russian.
????Wow! - she said in Russian.
288 Interjections across cultures
The reason why any combination of direct discourse with a reference
to translating is a little odd, has been suggested earlier. The reason why
primary interjections sound particularly odd in such circumstances has to
do, I think, with their peculiar semantic structure. The direct discourse
suggests that what is being reported is not only the content (the what)
of the original utterance but also the how. The phrase 'in language L'
indicates that in this particular case the how refers not to the actual
words but, presumably, to the illocutionary force of the original utter-
ance. But perhaps primary interjections do not have any illocutionary
force which could be separated from their actual form.
Haiman (1989:156) points out the deep semiotic difference between
utterances such as (a) and (b) below:
a. YukI
b. I feel disgust.
and (endorsing my own analysis in Wierzbicka 1973) he links the first
type with BUhler's (1933) Ausdrucksfunktion (expressive function of
language), and the second, with BUhler's Darstellungsfunktion (symbolic
or representational function of language). The essence of this difference
could perhaps be represented along the following lines:
7
YukI =
I feel disgusted
I feel disgusted! =
I say: 1 feel disgusted
I say this because I want to say what I feel
Similarly:
Ow! (or: Ouch!) =
I feel pain
I feel pain =
I say: 1 feel pain
I say this because I want to say what I feel
If these partial explications are valid, primary interjections have
no illocutionary force at all, because they include neither an 'I say'
component nor an illocutionary purpose (an 'I say this because ... '
component). Nonetheless, they do have a meaning. Having no illocution-
ary components (no components in the frame 'I say ... ') they are
Preliminary discussion 289
not speech acts. Rather, they are what one might call vocal gestures.
(Cf. Goffman 1981:78-123.)
But they are neither universal nor meaningless. On the contrary, they
are language-specific, and they are meaningful. Furthermore, we don't
have to be content to hint vaguely at what the meaning of a given
interjection might be, as dictionaries usually do, for example: "wow -
interjection used to express strong feeling (e.g. pleasure or surprise)" or
"oops - interjection used to express mild apology, esp. for carelessness"
or "ah - interjection used to express surprise, triumph, derision or
amused discovery" (LDOTEL 1984). They have a semantic invariant,
and this invariant can be revealed and formulated in such a way as to
account precisely for their range of use. This can be done without relying
on language-specific terms. such as 'disgust' (for yuk), 'impressed' (for
wow) or 'pain' (for ouch), which have to be explicated themselves if they
are to be meaningfully compared with emotion terms available in other
languages, and which in any case don't correspond exactly to the
meaning of any interjections. We can capture the subtlest shades of
meaning encoded in interjections relying exclusively on universal or
near-universal concepts such as 'good' and 'bad', 'do' and 'happen',
'want', 'know', 'say', or 'think'. (Cf. Wierzbicka 1972, 1980, 1985c,
1987, 1988.)8 For example, the English interjections oops and wow can
be explicated as follows (cf. Goffman 1981:102 and 108):
oops
1 now know: 9
1 did something bad
something bad happened because of that
1 didn't want it (to happen)
I would not want someone to think that it is very bad
(I feel something because of that)
wow
1 now know something
1 wouldn't have thought 1 would know it
I think: it is very good
lO
(I wouldn't have thought it could be like that)
1 feel something because of that
290 Interjections across cultures
1.2. Defining the concept of 'interjection'
An interjection can be defined as a linguistic sign (1) which can be used
on its own, (2) which expresses a specifiable meaning, (3) which does
not include other signs (with a specifiable meaning), (4) which is not
homophonous with another lexical item that would be perceived as
semantically related to it, and (5) which refers to the speaker's current
mental state or mental act (for example I feel ... , I want ... , I think ... ,
I know ... ) .
By these criteria, exclamations such as Good Lord!, Good heavens!,
Christ! or Hell! are not interjections, whereas those like gee, wow, oops
or ha are. (Cf. Goffman' s 1981:99 definition of what he calls "response
cries": "exclamatory interjections which are not full-fledged words".)
One class of signs for which the definition proposed above is some-
what problematic includes onomotopoeic signs such as hyc (jump) or
(fall) in Polish, which are meant to depict an action in a semi-iconic way
and which are used as substitutes for predicates, for example:
Kot hyc! - Z okna na (Grodzienska 1970:16)
'The cat hyc (i.e. jumped) - from the window onto the floor.'
Jak to slonko ujrzalo,
Tak glosno zasmialo,
Tak pod boki,
Ai - z chmury wysokiej.
'When the sun saw that,
It burst out laughing,
And was so delighted,
It [fell] - crunch! - from its place in the clouds.'
Iconic 'depictives' such as hyc or are somewhat problematic be-
cause in a sense they can occur as constituents of larger constructions
and because they are often homophonous with global signs to which they
are semantically related. For example, the Polish word hop can be used
either as a depictive representing a jump (and stressing, unlike hyc, its
vertical aspect), or as a global utterance urging the addressee to jump. If
we wish to categorise the urging hop as an interjection, its relationship
with the depictive hop presents a problem, in view of condition (4)
proposed above. If we lift this condition, however, we will be no longer
able to distinguish between 'primary' exclamations such as wow or gee
and 'secondary' exclamations such as Christ! or Hell!
Preliminary discussion 291
A possible solution to this difficulty would be to rephrase condition
(4) in such a way as to allow for homophony with certain categories of
signs (such as, for example, depictives, or particles) but not with others
(such as nouns, verbs, or adjectives). Tentatively, we might propose the
following: (4) which is not homophonous with another lexical item
whose meaning would be included in its own meaning (that is, in the
meaning of the putative interjection). If we adopted this phrasing of
condition (4), we would have to conclude that exclamations such as
Christ!, Hell! or Damn! mayor may not tum out to be interjections,
depending on what turns out to be a justifiable semantic formula for a
given exclamation. For example, if it can be shown that a semantic
formula for the exclamation Hell! should include the noun hell, then
this exclamation would not be categorised as an interjection; but if its
explication should not include the noun hell, then it would be so catego-
rised. This does not seem to be an unacceptable conclusion.
1.3. Types of interjections
If interjections have to refer to the speaker's mental state, or mental act,
then they can be classified on the basis of the exact nature of that state
or act. Thus, we could establish the following classes of interjections: (1)
emotive ones (those which have in their meaning the component 'I feel
something'); (2) volitive ones (those which have in their meaning the
component 'I want something' and which do not have the component 'I
feel something'; (3) cognitive ones (those which have in their meaning
the component 'I think something' or 'I know something' and which
have neither the emotive component 'I feel something' nor the volitive
component 'I want something'.
Other classifications of interjections can of course be proposed, based
on different criteria. For the present purposes, however, the division of
interjections into emotive, volitive and cognitive ones is useful, and the
following discussion will be based, essentially, on this division. It must
be pointed out, however, that this division cannot be adhered to very
strictly because of close semantic links between interjections of differ-
ent types. In particular, it is often the case that a cognitive interjection
has a homophonous emotive one, in which case a comparison of the two
meanings served by the same form is clearly called for.
292 Interjections across cultures
2. Volitive interjections
In Polish (and probably in many other languages as well), volitive inter-
jections can be divided into two classes of unequal importance: a minor
class of interjections directed at animals, and a major class of interjec-
tions directed at human beings. I will start with the first. (In the rest of
this chapter, interjections are cited without exclamation marks, even
though many of them have an obligatory emotive intonation.)
2.1. Interjections directed at animals
In Polish, reduplicated expressions such as cip-cip-cip, kici-kici-kici or
tas-tas-tas can all be assigned the same basic semantic formula: 'I want
you to come here'. The differences between them consist only in the
class of addressees:
cip-cip-cip
kici-kici-kici
tas-tas-tas
I want you (chickens) to come here
I want you (cat) to come here
I want you (rabbits) to come here
There is also at least one interjection (directed at non-humans) with the
opposite kind of meaning: 'I want you to go away from here'. The word
in question is sio (or a sio), and it corresponds rather closely to the
English shoo. Usually sio, like shoo, is directed at flies, or at birds,
and its meaning can perhaps be represented as 'I want you to flyaway
from here'.
Some interjections with astonishingly specific meanings are aimed
at horses:
wio 1 want you (horse/s) to move forward (cf. English gee-up)
prr 1 want you (horse/s) to stop
hejta 1 want you (horse/s) to move to the right
wista I want you (horse/s) to move to the left
Poland's hunting traditions are reflected in an interjection aimed at dogs
(and, by metaphorical extension, applied also sometimes to people):
huzia 1 want you (dog) to do something bad to this (these)
creature(s) (cf. English sic-em)
Volitive interjections 293
Volitive interjections directed at animals are no doubt very common
in different languages of the world. For example, Karcevski (1969: 198)
mentions that "to stop a horse both the Russians and the Finns shout
'tpru', that is to say, a long bilabial r". Goddard (1987) provides some
examples from the Australian Aboriginal language Pitjantjatjara:
t}u-t}uu
wii
'here boy' (a call to dogs)
'whoa!' (command to horse or command to pull up)
2.2. Interjections directed at people
Nearly all items in this class can be assigned the basic meaning: 'I want
(don't want) you to do something now'. Each item, however, contains
in addition some more specific components. The whole class can be
divided (somewhat arbitrarily) into the following subclasses:
1. The 'I want silence' group (sza, pst, cii);
2. The 'I don't want you in this place' group (won, precz, sio,
wara);
3. The 'I want you to jump' group (hop, hopla);
4. The 'urging' group (nuie, he), he}ie);
5. The 'communication over distance' group (hop hop, hallo,
ahoj);
6. The 'I give it to you' group (na).
I will briefly discuss some of these groups, trying to capture the invariant
of each interjection and illustrating it with some examples. In some
cases, 1 will compare the Polish interjections with their nearest (phonol-
ogical or semantic) counterparts in English or in Russian.
2.2.1. The 'I want silence' group
Polish has at least three different interjections aimed, roughly speaking,
at imposing silence: sza [Sn], pst [pst] and cii [tci:].
Sza (probably derived from cisza 'silence', as well as motivated by
sound symbolism) can be used, for example, when one has put one's
children to bed and when one wants to warn them not to talk any more,
whether aloud or in whispers, after one has left the room. There is a
294 Interjections across cultures
well-known saying urging children to be quiet: Cicho, dzieci, cicho sza
'quiet, children, quiet, sza'. (Cf. the English shush.)
Pst is somewhat conspiratorial. It could be said at a meeting or in a
classroom to someone who is whispering and who doesn't want to be
overheard by someone other than the intended addressee. The interjec-
tion seeks to warn the speaker that he can be overheard. It could also be
used when a secret is being discussed (for example a present which is
meant to be a surprise) and when the person approaches who is not
suppose'd to hear about it.
Cii! (derived no doubt from cicho 'quiet', but also perceived as
motivated by sound symbolism) can be said, for example, to one's
family, if one is trying to talk on the telephone and they are conducting
a conversation, or making some other noise, in the same room. But pst
or sza would never be used in such a situation.
The crucial differences are these. A person who says either pst or cii
tries to be quiet himself, whereas a person who says sza seeks to impose
silence on others but not on himself. Thus, one can call or shout sza,
but one cannot call or shout pst or cii:
Sza! sza! - zawolal na dzieci lyikami w pusty,
blaszany rondel. (Reymont) (SIP)
'Sza! sza!, he called at the crowd of children drumming their
spoons on an empty tin pan.'
This may be related to the fact that (as noted by Rosten 1968:327 with
respect to the homophonous Yiddish shah, no doubt related to the Polish
sza) sza is normally directed at a group of people rather than at an
individual person. But pst and cii can have, and perhaps usually have,
an individual addressee.
A person who says cii tries to prevent any noise (it could be whispers,
but it could also be footsteps, or some rustling noise); and it is presup-
posed that some effort directed at not being heard is already being
made. A person who says pst tries to prevent audible speech, rather
than any other kind of noise; and may be seeking to impose silence on
some particular topic (for example a surprise party) rather than absolute
silence, as cii does.
The different though related meanings of the three interjections in
question can be represented as follows:
sza
I don't want anyone of you to say anything now
Volitive interjections 295
I don't want people to hear you now
you know you have to do what I say
pst (with a finger on one's mouth)
I don't want you to say anything now
I think someone other than I could hear you say it
I don't want someone other than me to hear you say it
I don't want someone other than you to hear me say this
cii
I don't want you to say or do anything now that people could
hear
I think they could hear you now
I don't want someone other than you to hear me say this
Sza, pronounced fB, resembles of course, both in form and in meaning,
the English interjection sh, "often used", according to LDOCE (1978),
"in prolonged or reduplicated form, to urge or command silence". As its
different form suggests, however, the English sh is 'quieter' than the
Polish sza: sza can be said quite loudly itself, whereas the English sh
cannot be said very loudly, and seeks to include the speaker in the
imposition of silence. The explications of sza and sh can therefore be
differentiated as follows:
sza =>
I don't want people to hear you now
sh =>
I don't want people to hear anything here now
The Polish pst brings to mind, of course, the English psst. But here,
too, there is a difference in meaning, if not in form. The English psst
suggests that the speaker wants to say something to the addressee
(unbeknown to other people). For example, in English one might say psst
to draw somebody's attention to a beautiful bird, without frightening
the bird. But the Polish pst would never be used like that. It could be
used to prevent the other person from speaking (in a way that the bird
could hear), but not to draw that person's attention to the bird. Accord-
ingly, the explication of the English psst should include one additional
component: 'I want to say something to you now':
psst (English)
I want to say something to you now
296 Interjections across cultures
1 don't want someone other than you to hear it
I don't want someone other than you to hear me say this
As these explications show, both the English psst and the Polish pst have
a 'conspiratorial' quality; but the English psst indicates that the speaker
wants to say something (in a private and 'conspiratorial' manner),
whereas the Polish pst indicates that the speaker wants to prevent the
addressee from saying something that an outsider could hear.
A similar interjection (pst, pss or psst) exists also in Russian, where it
is used, however, quite differently. The Academy dictionary of Russian
(SSRLJa 1950-65) defines it as follows: "it is an onomatopoetic interjec-
tion used to express displeasure, disapproval, warning etc." For example:
Mne ne ponravilsja glavnokomandujuseij - meloenyj celovek,
glupyj ... Siroty - nul' ... Johann Hensen proiznes prezritel' no -
Psst! (A. Tolstoy) (SSRLJa)
'I was not impressed by the commander-in-chief - a petty man,
stupid ... No breadth whatsoever! ... Johann Hensen uttered
contemptuously: Psst! '
Clearly, the English and the Polish pst have nothing to do with express-
ing displeasure or disapproval, and at the level of semantic analysis
at which one operates with global terms of this kind the Russian pst
must be seen as entirely unrelated to the English or Polish one. If,
however, we attempt to capture the invariant of the Russian pst in terms
of components formulated in the universal semantic metalanguage we
do discover some shared elements of meaning:
pst (Russian)
1 think something bad about this
I don't want to say now other things about it
(I feel something bad thinking about it)
It appears that despite the considerable differences between the Polish,
the English, and the Russian pst, they all can be assigned the component
'I don't want (you) to say something'.
2.2.2. The 'I don't want you in this place' group
Polish has at least three different interjections (directed at people) with
the general meaning 'I don't want you to be here' or 'I don't want you to
be in this place': won, precz and wara.
Volitive interjections 297
Won and precz both imply 'bad feelings', but won is also contemptu-
ous ('I think something bad about you, you are like an animal'). Precz
can be used as a semi-adverb as well as an interjection (idi precz! 'go
away! ') but won is normally used only as an interjection; presumably,
its contemptuous character constitutes an obstacle to its use as an
adverb. As for precz, it usually conveys hostility (though not contempt),
but it can also be used with 'bad feeling' aimed at something other
than the addressee, as in Adam Mickiewicz's poem addressed to his
lost love:
Precz z mej p m i ~ c i
'Go away from my memory!'
It is inconceivable that won could be substituted here for precz.
As for wara, it is related to the verb warczec, 'growl', and it evokes
the image of a growling dog, showing a bad feeling and warning the
addressee to stay away from a place. Examples for won and wara are:
Ryknfl/ basem - Won! - potem jeszcze raz, glosniej - Poszed/
won! (Zukrowski) (SIP)
'He roared: - Won! (Get out of here!), and then once again,
louder 'Be gone, Won!'
Koziel, kt6ry tam wlasnie przyszedl wody szukac:
Ej - krzyknfll z g6ry - Ej, ty ryzy kudla, wara od ir6dla.
I hop w d61. (Mickiewicz) (SIP)
'A billy-goat that had come in search of water shouted: Ej, ej,
you gingertail, keep away (wara) from the spring. And hop off
down the hill.' [Note the use of wara, ej and hop.]
The three interjections may be explicated as follows:
precz ('go away')
I don't want you to be here
I want you to go away from here (now)
I feel something bad thinking about it
won ('get out!') (cf. English g'wan)
I don't want you to be here
I want you to go away from here (now)
I think something bad about you (you are like an animal)
I feel something bad towards you
298 Interjections across cultures
wara (gesturing with a finger)
I don't want you to come to this/that place (where X is)
I feel something bad thinking about it
I will do something bad towards you if you do it
2.2.3. The 'I want you to jump' group
Both hop and hopla can be used not only as volitive interjections but
also as expressions accompanying a jump and, ostensibly, depicting it
in what is perceived as an onomatopoeic manner (as in the example from
Adam Mickiewicz's poem, quoted under wara).
hop
I want you to jump (now)
(or: to move quickly like someone who jumps)
hopla
I want you to jump over this thing (now)
2.2.4. The 'urging' group
Nuie is usually used with an imperative or an infinitive (in the function
of a command), but it can also be used on its own, or with a form of
address. It indicates that the addressee should hurry. For example:
Nuie, nuie! Wstawaj, dose spoczywac! (Gruszecki) (SIP)
'Come on, come on! Up you get, that's enough lazing about!'
Nuie szelmy, jese mu dae, nim pacierz minie, bo lby pourywam!
(Sienkiewicz) (SIP)
'Come on you scoundrels, give him some food in a wink, or I'll
pull your heads off.'
Its meaning can be represented as follows:
nuie
I want you to do it now
I don't want you to do it after now
Hej is also trying to mobilise the addressee to action, but it doesn't
imply that the addressee is (or can be expected to be) slow, and it has an
Volitive interjections 299
additional function of an attention-getter. Thus, trying to urge the
addressee to get up quickly, one would say nuie rather than hej; but
calling to someone who is some distance away and who doesn't know
that he is being addressed, one would say hej! rather than nuie. Some
examples:
Hej! z drogi! ... wolal na kupig,cych ludzi. (Sieroszewski)
(SIP)
'Hej! Off the road! ... he shouted at the gathering of people.'
He}! he}! czlowieku! a gdzie to idziecie? Zblfldziliscie - nie
droga. (Zmorski) (SIP)
'Hej! hej! You there! Where are you going? You've taken a
wrong turn. That's not the way.'
He}, szynkarko, w6dki, w6dki! (Krasicki) (SIP)
'He}, waitress, vodka, vodka!'
The meaning of hej can be represented as follows:
hej
I want to say something
I want you to hear it
I want you to do something now
He} has also another use as an emotive interjection; and the two uses
are closely related, but this point will not be discussed here.
Hejie (formally, a combination of hej and the particle ie) urges the
addressee to do something quickly, but it adds to this volitive aspect
a component of 'merriness' and it implies an invitation to a 'merry
action'. It is not clear whether it carries the other ('attention-getting')
components of he}. For example:
Hejie do roboty, bierzcie iywo mloty. (Lenartowicz) (SIP)
'Hejie let's get to work, grab your hammers.'
'He}ie! w czasze! (Zielinski) (SIP)
'He}ie! Drain your glasses!'
The meaning of he}ie may be represented as follows:
hejie
I want you to do something now
I think you will feel something good when you are doing it
300 Interjections across cultures
2.2.5. The 'communication over distance' group
In this group, the most interesting item is hop-hop, which is used to
call people loudly, over considerable distance (especially in a forest,
but not necessarily so). For example:
Hop, hop! ... hop! hop! ... - rozlegalo ~ po lesie. Wiele razy
powtarzaly ~ te okrzyki bez odpowiedzi. (Skiba) (SJP)
'Hop, hop! ... - there resounded a cheerful call from the direc-
tion of the tower.'
Hop, hop, panie, ezy tu mieszka pan Grelowicz? (Breza) (SJP)
'Hop, hop, does Mr Grelowicz live here?'
The meaning of the Polish hop hop can be portrayed as follows:
hop hop
I know that you are far away from me
I want you to hear me
I want you to say something
What is particularly interesting about this interjection is its relation-
ship to comparable interjections in other languages. Thus, the Russian
au [au] appears to have a somewhat different range of use, stressing
visibility and contact rather than distance. For example, one can call au
when entering a friend's flat or house (if the friend cannot be seen); but
one would never say hop hop in this situation. SRJa defines au as "a call
that people use in order to find, or not to lose, one another", and SSRLJa
defines it as a "reciprocal call used by people who can't see one an-
other". For example:
Kogda odna uxodila v sad, to drugaja uze stojala na terrase i,
gljadja na derev'ja, oklikala: "au, Zenja!" iii "Mamocka, gde
ty?" (Cexov) (SSRLJa)
'When one went off into the wood, the other would always
stand on the porch, gaze at the trees and callout: "au, Zenja!" or
"Mummy, where are you?'"
Masa: fa pojdu poiscu ego.
Arkadina: Pozalujsta, milaja.
[Masa (idet v levo)}: Au! Konstantin Gavrilovic! ... Au! (Cexov)
(SRJa)
'Masa: I'll go and look for him.
Volitive interjections 301
Arkadina: Please do, my dear.
[Masa (goes to the door)] : Au! Konstantin Gavrilovic! ... Au!'
The combination "Au! Konstantin Gavrilovic!" suggests that one can
call au! without assuming that one is very far from the target person.
But in Polish, hop hop does imply the assumption that the target person
is far away and can only be reached by shouting. Furthermore, au
implies that the speaker doesn't know where the addressee is, but this
is not necessarily the case with hop hop (cf. the last example of hop
hop where the speaker and the addressee can in fact see each other).
It is also interesting to compare the Polish hop hop and the Russian
au with the Australian English interjection cooee (apparently from an
Aboriginal word imitating the sound of a bird), used by people to com-
municate over distance in the bush. Cooee is similar to au (and different
from hop hop) in its stress on visibility, as it could not be used by
people who see one another. It is different from au, however, in so far
as it implies that both the speaker and the addressee don't know where
the other person is (since they are both walking in the bush); whereas in
the case of au, the addressee may know where the speaker is (as in
the first Chekhov example above). Finally, au could be uttered when
entering a friend's house (if the house is large and if one thinks that the
friend is somewhere in the house but cannot see him or her); but one
could never use cooee or hop hop like that (except in jest).
cooee
I know we are now in a kind of place where people can't see one
another (if they are not in the same part of that place)
I can't see you
I think you are far away
I want to know where you are
I want you to know where I am
I say this in this way because I want you to hear me
au
I can't see you
I want to know where you are
I want you to hear me
302 Interjections across cultures
3. Emotive interjections
3.1. Interjections of 'disgust' and similar feelings
Haiman (1989: 157) calls the English interjection yuk "a language-
specific verbalisation ... of a universal gesture of revulsion". He sup-
ports his contention that yuk is a "language-specific verbalisation" with a
reference to the German interjection pfui and to the Dakota interjection
xox. Apparently, however, he sees only the form of these interjections
as language-specific. Their meaning is seen in terms of "a universal
gesture of revulsion".
But is this meaning really universal? Since emotive interjections are
less conventional than lexical terms of emotion, it is perhaps natural to
expect that their meanings may be less language-specific than terms
such as the English anxiety, the German Schadenfreude, the Polish
tesknota or the Russian xandra. Nevertheless, it is an illusion, I think, to
believe that there is for example 'a universal gesture of revulsion'. For
example, Polish does not have an exact equivalent of yuk. On the other
hand, it has three other interjections: fu, fe, and tfu, which could be
linked, loosely speaking, with something like disgust, but which none-
theless differ from one another and differ from yuk.
3.1.1. The Polishfu and the English yuk
The Polish interjection fu has a homophonous counterpart in Russian,
but the meaning of these two fus is not the same. Karcevski (1969: 198)
glosses the Russian fu as an interjection of "repulsion oifactive", that
is, 'olfactory revulsion'. This description (whose adequacy with respect
to Russian we will consider later) would fit the Polish interjection fu
to some extent, as it is quite likely to be used when one is suddenly
confronted with a bad smell. However, it can also be used in other
circumstances, as for example when one sees a person licking off some-
body else's plate. One might also say fu when one discovers in one's
refrigerator some item of decaying food, or when one is invited for the
first time to eat snails. On the other hand, one would not say fu if a
repulsive creature such as a spitfire caterpillar or some particularly
repellent worm landed on one's arm, or when one saw a squashed slug
on the footpath - both situations when one could easily say yuk.
Emotive interjections 303
The generalisation seems to be this: one says fu when confronted with
a 'disgusting' smell, 'disgusting' food or with 'disgusting' eating habits,
that is, when one's nose or mouth comes close to something perceived
as 'bad' and when one feels something bad because of that and wants
to avoid the offensive contact (or closeness).
But if a worm landed on one's nose one would not say fu, presumably
because the bad feeling caused by this contact is only accidentally
linked with the nose (having a repulsive worm land on one's nose does
not seem different from having it land on one's cheek). Fu seems to be
restricted to feelings which are necessarily defined in terms of nose or
mouth - that is, to situations of introducing something 'repulsive' into
one's nose or one's mouth.
To account for the range of use of the Polish fu, I would propose (as
a first approximation) the following explication:
fu (Polish)
I now know something about something in this place
I feel something bad in my body because of that
I feel like someone who thinks: I don't want this thing to come to
be in my nose or in my mouth
I think other people would feel the same
The component 'I think that other people would feel the same' is
intended to capture the speaker's notion that there is something 'objec-
tively' disgusting or repugnant about the situation. The subcomponent
'in my body' is intended to show that this interjection suggests a physi-
cal reaction on the part of the speaker, and not, for example, moral,
intellectual or aesthetic disgust.
In the case of yuk, the prototype is no doubt physical, too. LDOTEL,
which describes yuk (yuck) as an "interjection of disgust, probably
imitation of the noise of retching", may well be right in this respect.
Nonetheless many of my informants (especially younger ones) affirm
that they use this word not only to express a physical reaction (to a
physically disgusting or revolting object), but also in the presence of
a non-physical stimulus; for example, some students say that they
would say yuki when given by their lecturer an assignment which they
strongly dislike.
II
This disagreement among informants may reflect
some sociolinguistic variation: the use of yuki is spreading and its range
of use appears to be widening. The following quote from The Australian
newspaper (25.10.88:14) shows that - at least in Australia - yuk in an
abstract (non-physical) sense can now appear in writing and even in
304 Interjections across cultures
print: "FEMINISTS - YUK! As a woman I feel ashamed" (from a letter
to the Editor expressing the reader's reaction to new guidelines for
gender-inclusive language, called on the same page "Sheila-Speak"). In
the case of the Polishfu the presence of a 'bodily component' is indubi-
table. For example, one could never say fu in response to some new set
of guidelines, or to an intellectual task such as a university assignment. 12
Yuk differs in this respect from another (perhaps older) English inter-
jection which is used to express disgust: phew (which may be segmen-
tally homophonous with the phew of relief mentioned earlier but which
has a different tone pattern, sometimes represented as pee-yew and the
like). The 'phew of disgust' is linked specifically with what Karcevski
called 'olfactory revulsion', and - like the German pfui mentioned by
Haiman (1989) - it refers clearly to a bad smell. It is nonetheless
perceived as semantically similar to yuk. For example, in an advertise-
ment for garbage bags frequently shown on Australian television, a
cuckoo jumping out of a clock and seeing a heap of garbage utters
consecutively phew and yuk.
Tentatively, I would propose for yuk the following explication:
yuk
I think: this is bad
I feel something bad because of that
I think other people would feel the same
I feel like someone who thinks:
I don't want to be in the same place as this
3.1.2. The Russianfu
If the Polish fu is similar to the English phew (the phew of disgust) in
being clearly related to bodily sensations, although not necessarily
olfactory ones, it differs in this respect from the Russian fu, which,
as mentioned earlier, was glossed by Karcevski (1969:198) as an inter-
jection of 'olfactory revulsion', but which in fact can express a purely
mental (moral, aesthetic etc.) reaction. Consider for example the follow-
ing passage:
... ej xotelos' pojti v spal' nju Verocki, sobrat'sja k nej pod ode-
jalo i prilaskat' ee ... No ona otgonjala eti uzasnye mysli. Fu!
kak uiasno! Kak kakaja-nibud' lesbijanka! Gadost' kakaja!
(Suslov 1982: 188)
Emotive interjections 305
' ... she was tempted to go to Verocka's bedroom, to sneak under
her covers and to cuddle her ... But she would chase such awful
thoughts away, in horror. Fu! How horrible! Like some lesbian!
How disgusting!'
One might say that in this case the speaker's disgust is almost physical,
but it is caused by perceiving 'disgusting' thoughts in one's head, not by
perceiving 'disgusting' smells or sights. The Polish fu could never be
used in a similar context, and the English yuk would be unlikely to be
used in this way.
In fact, the Russian fu can also be used in situations which are even
further removed from any physical disgust or revulsion. The Academy
dictionary of Russian (SSRLJa) glosses fu as an interjection of "con-
tempt, annoyance, revulsion (otvrascenie) etc.", mentioning revulsion
as only one of the possibilities, whereas Dal's dictionary (1882) doesn't
mention it at all, glossing fu as "an interjection of contempt or annoyance
(mezdometie prezrenija, dosady)". The examples cited in these dictionar-
ies refer to human behaviour evaluated as 'stupid', or to irritating events
concerning the speaker. For example:
Fu, kak vy bestolkovy! 00. Sejcas ze pozovite eOtix niscix! (Kuprin)
(SSRLJa)
'Fu, how stupid you are! ... Call these beggars in at once!'
Fu, kakoe skvernoe pero! - zakrical Sumnov, udariv v dosade im
po stolu. (Dostoevskij) (SSRLJa)
'Fu, what an awful pen! - shouted Sumnov, angrily hitting it
against the table.'
To account for the non-physical character of the Russian fu, we can
perhaps explicate it along the following lines:
fu (Russian)
I now know/imagine something
I think: this is bad
I feel something bad because of that
I think that other people would feel the same
I feel like someone who thinks: I don't want this
If we wanted to take into account Karcevski' s suggestion that the Russian
fu does have some link - however tenuous or metaphorical - with the
nose, we could rephrase the last component of this explication as
follows: 'I don't want this thing to be near my nose'; this would not
306 Interjections across cultures
be incompatible with the examples adduced above because moral or
aesthetic 'disgust' could be conceptualised in terms of a physical
prototype. It is important, however, not to include in this explication
the component 'in the body', which we have assigned earlier to the
Polish fu.
3.1.3. The Polishfe
Unlike the Polishfu, the Polishfe is similar to the Russianfu in not being
restricted to physical sensations, although it differs from it in other
respects. Roughly speaking, fe expresses a mild moral disgust. It is
always a reaction to human behaviour, and it could never be used, like
the Russian fu, with respect to an annoying object (for example a pen).
The feeling expressed is 'mild', or rather it is expressed in a 'mild'
manner, because the speaker's purpose seems to be didactic rather than
purely expressive: by expressing a negative emotional reaction to the
addressee's behaviour the speaker is trying to shame the addressee, and
to cause him to correct his behaviour.
Typically, fe is used by adults speaking to small children, but it can
also be used, for example, by a mild-mannered woman flirtatiously
scolding a man. SJP glosses fe as "an interjection expressing disgust,
displeasure, reprimand", but this somewhat heavy-handed gloss is
followed by a much more helpful comment: "it means something
similar to 'you should be ashamed of yourself! that's not nice! '''. Some
examples:
Rzepa takze go zagadnie:
Fe! Niedobrze! Fe! Nieladnie!
Jak pan moze,
Panie pomidorze? (Brzechwa 1983:23)
'The turnip will also reprove him:
Fe! That's bad! Fe! That's not nice!
How can you, Mr Tomato?'
Fe! Jak pan moie mowic takie rzeczy! (Sienkiewicz) (SJP)
'Fe! How can you (polite form) say such things!'
Fe, Sewerynku, wstydi s ~ zlym bratem jestes. (Bliziilski) (SJP)
'Fe, Jimmy (Sevie), you should be ashamed of yourself, you're
a bad brother.'
Emotive interjections 307
In fact, the most common collocation is fe, nietadnie 'fe, that's
not nice', as in the following characteristic examples from children's
poems:
Fe, nie/adnie! Fe, k/arnczucha! (Brzechwa 1983:9)
'Fe, that's not nice! Fe, you're a little liar!'
Fe, nieladnie! ktoz tak k/arnie?
Zaraz rnarnie! (Brzechwa 1983:9)
'Fe, that's not nice! Who lies like this?
I'll tell mum about it!'
Other common collocations involving fe are 'fe! how can you', or 'fe!
you should be ashamed of yourself!', as in the examples above. But
fe can also occur on its own, as in the following children's poem:
Brudasek
A ten piesek Brys,
co nie chcial kg,pac dzis,
rna na lapkach kurz i piasek,
spac pojdzie jak brudasek.
Fe! (Szelburg-Zarembina 1970:60)
'Dirty little grub.
That little dog Spottie,
who didn't want to have a bath today,
has sand and dust on his little paws,
so he'll go to sleep like a dirty little grub.
Fe!'
It is interesting to note in passing that there is a similar-sounding
interjection of disgust in Danish and Swedish (Anne Dineen and Jean
Harkins, p.c.), which has a similarly didactic purpose: fy, and which
appears to be used, prototypically, with reference to toilet training.
However, while a small child who has soiled himself may well be
reprimanded in Danish or Swedish with a disgusted fy, the Polish fe
would be much less likely to be used in such an elementary situation,
as it appeals to higher-level social and moral rules. One might say that
Ie appeals to the addressee's sense of shame, whereasfy seeks to activate
a Pavlovian reflex of elementary disgust.
fy
I now know that you have done something bad
308 Interjections across cultures
1 feel something bad because of that
I think that other people would feel the same
I want you to feel something bad because of that
I don't want you to do this
fe
I now know that you have done something bad
I feel something thinking about it
I think that other people would say the same
I think you know that if someone does something like this it
is bad
I want you to feel something bad thinking about it
Fe is a somewhat genteel or 'refined' thing to say, which appeals to the
addressee's sense of shame and knowledge of social rules, suppressing,
as it were, an impulse of disgust. For this reason, I have phrased the
feeling itself more vaguely as 'I feel something thinking about it' rather
than as 'I feel something bad because of that'. (For further discussion of
this point, see section 3.1.6.) On the other hand, I have included in the
explication of fe a sentence referring to the addressee's understanding of
certain general rules of civilised behaviour ('I think you know that if
someone does this it is bad'), which 1 have not included in the definition
of the more elementary fy.
3.1.4. The Yiddishfeh
In a Russian-Jewish popular song, called 'A Soviet Jew's prayer' or
'A Jewish marseillaise', the Yiddish interjection fe is used as an overt
signal of Jewishness:
Otrecemsja ot starogo mira ...
Nam ne nuzno zlatogo kumira - Fe!
'Let's renounce the old world ...
We don't need a golden calf - Fe!'
Fe represents here the Soviet Jews' mock-identification with 'Soviet'
values and their mock-rejection of, and mock-disgust for, any other val-
ues (in particular, capitalist values).
According to Rosten (1968:115), "Feh! is the Yiddish replacement
for exclamatory expressions of disgust such as 'Phew!', 'Pee-oo!',
'U h" 'Ph I' 'E h" d P ~ h I' I . 'F h"
g ., ooey., cc. an Jrr..... n sayIng e., you may
Emotive interjections 309
bare the teeth and wrinkle the nose, in visible reinforcement of the
meaning". Rosten says that he "once wrote an entire story to illustrate
the puissance of this incomparable expletive", and he lists (among
others) the following situations "in which Feh! may serve as the perfect
utterance":
1. Smelling a rotten egg.
2. Passing an open sewer.
3. Inhaling Los Angeles smog.
4. Driving past the sulfur pits that fringe New York in New
Jersey.
5. Whiffing a rotten fish.
6. Depicting a beatnik with mare's-nest hair.
7. Summarising a political position you detest.
8. Appraising the honor or benevolence of an enemy.
9. Contemplating an operation for haemorrhoids.
10. Delineating the character of the paskudnyak who ran off
with your wife.
11. Recounting how a soprano murdered an aria.
12. Depicting a hangover.
It seems likely that the Yiddish feh comes from the Polish fe (or
possibly the other way around). But it is clear that the range of use of the
Yiddishfeh is much wider, and that in fact it is at least as wide as that of
the English yuk. It is not, however, as wide as the last point on Rosten's
list would suggest: "portraying strongly negative feelings about any
sight, event, person, crisis, experience or emotion". As Rosten is well
aware, Yiddish has some other interjections expressing "strongly
negative feelings", in particular oy vay, for which feh is by no means
always substitutable.
The crucial difference between these two interjections seems to lie in
a more 'disinterested' and more 'descriptive' character of feh. Oy vay
implies: 'something bad is happening to me'; feh implies: 'this is bad' -
without the element of personal involvement 'to me' and without the
dynamic element 'is happening'. But feh is not merely lacking in
personal involvement: it positively claims that something is bad
'objectively', and that other people, too, would feel something bad
because of that. For example, if a brick fell from a roof and hit one on
the head, one could not exclaim Feh! To account for this 'public' charac-
ter of feh I would posit for it, once again, the component 'I think that
310 Interjections across cultures
other people would say the same'. This leads to the following partial
explication of feh (cf. section 3.1.6):
feh
1 think: this is bad
1 feel something bad because of that
1 think other people would say the same
1 think other people would feel the same
In view of the wide range of use of feh, I haven't included in this
explication a reference to the speaker's body. It is interesting, however,
to note Rosten's observation on the 'wrinkling of the nose' and the
'baring of one's teeth', which he thinks reinforce the meaning of feh. I
presume that these gestures indicate a mild version of olfactory and oral
repugnance: the speaker behaves as if he was trying to minimise the
contact of his nose and his lips with some offensive substance. These
gestures of repugnance are not as violent, however, as to suggest an
image of a person trying to get rid of some such substance from their
nose, mouth or throat.
Leaving aside the question of any references to body parts and bodily
gestures, the formula proposed above for feh is virtually identical with
that proposed earlier for yuk - except for the component 'I think that
other people would say the same', which presents feh as being more of a
judgement than yuk is, and which links it with the highly judgemental
Polish fee Both feh and yuk involve a judgement and a gut reaction, to
some extent, but in yuk the latter element appears to be stronger, and
it seems more marked in abstract contexts (such as, for example,
recounting how a soprano murdered an aria) than in the case of feh.
3.1.5. The Polish tfu and the Russian t'fu
Turning now to the Polish interjection tfu and to the Russian interjection
t'fu, we must note, first of all, that they both indicate the action of
spitting. Karcevski makes this point about the Russian t'fu, which he
links with the Russian expression naplevat' (lit. 'to spit'), and which he
also compares with the German expression Ich spucke drauf! 'I spit on
this'. As pointed out by Karcevski, the Russian t'fu can be used as a
substitute for naplevat', and can function as the main verb.
Mne
to-me
=
na
on
eto
this:ACC
Emotive interjections 311
nap/evat' !
to.spit
fa t'fu na eto!
I:NOM t'fu on that:ACC
'I spit on this (that's how I feel about it)!'
But in Russian, naplevat' and, consequently, t'fu, means more than an
outsider could infer from the information that both these words sym-
bolise the action of spitting: spitting itself is a symbolic action with its
own culture-specific meaning. In Russian, naplevat' '(I can) spit (on
that)' expresses a contemptuous indifference. 'I don't care', the speaker
seems to be saying, adding to this a defiantly coarse touch. T'fu as a
substitute for naplevat' can mean the same. For example:
A naplevat' mne na nego, on mne nipocem! (SSRLJa)
'I can spit on him, I don't need him!'
Mne na eto vase zoloto - t'fu! I ona dejstivitel' no pljunula sebe
pod nogi. (Polevoj) (SSRLJa)
'I can t'fu (spit) at that gold of yours! And she actually spat in
front of her feet.'
This sense of t'fu (and of naplevat') can be explicated as follows:
t'fu (Russian)
I don't care about this
I feel something bad thinking about it
I feel like someone who wants to do this: [spit]
The expression I don't care used in this explication can be further
explicated along the following lines:
I don't want to think about this
I don't want someone to think that I want this
The link between the symbolic act of spitting and an attitude of con-
temptuous indifference is particularly clear in the following passages:
- Pojdem, - govorit, - posmotrim tvoi ramy.
- U menja, - govorju, - urok dolzen byt'.
- Da pljun' ty, - govorit, na urok, raz takoe delo.
(Goljavkin 1968:110)
"'Let's go", he says, "and have a look at your frames."
312 Interjections across cultures
"I can't", I say, "I am due for a lesson."
"Stuff the lesson" [lit. 'spit on the lesson'], he says.'
SSRLJa posits also another meaning of t'fu, asserting that "it is also used
to express displeasure, annoyance, disappointment, etc." This alleged
second meaning is illustrated with examples such as the following one:
Zato u madam Bubnovoj ... - T'fu ty so svoej Bubnovoj! -
Aleksandra Semenovna vybezala v velicajsem negodovanii.
(Dostoevskij) (SSRLJa)
'On the other hand, as for Mrs. Bubnova ... "T'fu [I spit] on you
and your Mrs. Bubnova!" and Aleksandra Semenovna ran out
extremely indignant.'
It appears, however, that examples of this kind fit very well the formula
postulated for naplevat' and for the corresponding use of t'fu, and that
the postulated polysemy of t'fu is not justified.
On the other hand, the Polish tfu is used differently, although it, too,
symbolises the action of spitting. Roughly speaking, it expresses not
personal indifference (contemptuous indifference), but contempt and
moral disgust. SJP cites (among others) the following examples:
Tfu! wstyd! warcholilismy ~ okrutnie, a teraz trzeba przed cnotfl
i niewinnoscifl swiecic. (Sienkiewicz) (SJP)
'Tfu! shame on us! we behaved like oafs and rascals, and now we
have to answer for our behaviour before the virtuous and the
innocent! '
Tfu! Mospanie Hrabia, czy Wasc zb6jca? (Mickiewicz) (SJP)
'Tfu! Count, are you a robber?'
The meaning encoded in the Polish tfu appears to be this:
tfu (Polish)
I think: X did something bad
I wouldn't have thought that someone like X could do that
I feel something bad thinking about it
I think that other people would feel the same
(I feel like someone who wants to do this: [spit] )
It is interesting in this connection to recall Darwin's comments on the
meaning of spitting:
Spitting seems an almost universal sign of contempt or disgust; and
spitting obviously represents the rejection of anything offensive from the
Emotive interjections 313
mouth. Shakespeare makes the Duke of Norfolk say, 'I spit at him - call
him a slanderous coward and a villain'. So, again, Falstaff says, 'Tell
thee what, Hal, - if 1 tell thee a lie, spit in my face'. Leichhardt remarks
that the Australians 'interrupted their speeches by spitting, and uttering a
noise like pooh! pooh! apparently expressive of their disgust'. And
Captain Burton speaks of certain negroes 'spitting with disgust upon the
ground'. Captain Speedy informs me that this is likewise the case with
the Abyssinians. Mr. Geach says that with the Malays of Malacca the
expression of disgust 'answers to spitting from the mouth'; and with the
Fuegians, according to Bridges, 'to spit at one is the highest mark of
contempt'. (Darwin 1955:260)
The examples adduced, however, show that spitting does not have the
same meaning in different cultures. Disgust, contempt, disdain, and
moral indignation do indeed have something in common, but the differ-
ences between them are no less real than the similarities.
3.1.6. 'Disgust' and bodily gestures
Concluding our discussion of selected interjections of 'disgust and
similar feelings' we will note that in most cases they appear to be
linked with certain physical gestures, and that this may well be the
reason why they can be perceived as 'natural' (that is, non-arbitrary).
Thus, the Polish tfu and the Russian t'fu are perceived as imitations
of an act of spitting (although their meaning is by no means the same).
The English yuk can be perceived as an imitation of the sound of
retching. The English phew, the German pfui and the Scandinavian fy
can be thought to imitate an attempt to breathe out of one's nose a
repulsive smell (although again, their meaning is by no means the same).
The Polish and Russianfu can be perceived as associated with 'olfactory
or oral revulsion', although the use of the Russian fu, in contrast to that
of the Polish one, is not restricted to physical sensations. The close
phonetic resemblance between fu and the interjections symbolising the
act of spitting (tfu and t'fu) supports, I think, the idea that the Polish
and Russian fu is associated with an olfactory/oral revulsion, rather
than with a purely olfactory one.
Despite the different range of use of the interjections under discussion
(which of course must be reflected in adequate definitions), one might
consider including in their definitions a physical prototype. This could
be done as follows:
314 Interjections across cultures
tfu, t'fu ('oral repugnance') =>
I feel like someone who thinks:
I don't want this to be in my mouth
I want it to go out of my mouth
phew, pfui ('nasal repugnance') =>
I feel like someone who thinks:
I don't want this to be in/near my nose
I want it to go away from my nose
fu, fy ('nasal/oral repugnance') =>
I feel like someone who thinks:
I don't want this to be in my nose or in my mouth
I want it to go away from my nose/mouth
yuk ('pharyngal repugnance') =>
I feel like someone who thinks:
I don't want this to be in my throat
I want it to go out of my throat
feh, fe ('nasal/labial repugnance') =>
I feel like someone who thinks:
I don't want my nose or my lips to be in the same place as this
It might be added that the Polish fe could also be thought of as a euphe-
mistic version of 'nasal/oral repugnance'. The consonant may be seen
as a symbol of something like repugnance whereas the vowel can be
perceived in terms of a 'deliberate' vowel change: the speaker wants
to express his (or, more typically, her) repugnance in a softened and
'genteel' fashion, that is to say, without really sounding like someone
who gives way to their bodily impulses. This could be represented as
follows:
fe =>
I feel like someone who thinks:
I don't want my nose or my lips to be in the same place as this
I don't want to say it like this
Components referring to specific body parts and bodily gestures would
facilitate, I think, an inquiry into the sound symbolism of interjections
in a cross-linguistic perspective. In view, however, of the exploratory
nature of the present chapter, I don't want to argue strongly for their
inclusion in the semantic formulae; rather, I wish at this stage to offer
Emotive interjections 315
them only for the reader's consideration, together with the following
quote from Darwin:
As the sensation of disgust primarily arises in connection with the act of
eating or tasting, it is natural that its expression should consist chiefly in
movements round the mouth. But as disgust also causes annoyance, it is
generally accompanied by a frown, and often by gestures as if to push
away or to guard oneself against the offensive object. ... With respect to
the face, moderate disgust is exhibited in various ways; by the mouth being
widely opened, as if to let an offensive morsel drop out; by spitting; by
blowing out of the protruded lips; or by a sound as of clearing the throat.
Such guttural sounds are written ach or ugh; and their utterance is some-
times accompanied by a shudder, the arms being pressed close to the sides
and the shoulders raised in the same manner as when horror is experienced.
Extreme disgust is expressed by movements round the mouth identical
with those preparatory to the act of vomiting. The mouth is opened widely,
with the upper lip strongly retracted, which wrinkles the sides of the
nose, and with the lower lip protruded and everted as much as possible.
This latter movement requires the contraction of the muscles which draw
downwards the corners of the mouth. (Darwin 1955:257-258)
3.1.7. 'Disgust' and sound symbolism
We have examined a number of interjections expressing, loosely speak-
ing, disgust or similar feelings, and we have seen that they are not
necessarily phonetically similar to one another. For example, the English
yuk is clearly not similar to the German pfui or to the English phew or
pooh. At the same time, it is easy to see that many - though not all -
interjections of 'disgust' and similar feelings are phonetically similar to
one another: they tend to consist of a bilabial or labiodental voiceless
fricative (or a bilabial plosive), followed by a close vowel. This particu-
lar phonetic structure can be understood as an icon of a 'natural' oral or
nasal gesture: an attempt to blow away a bad smell or to get rid of some
undesirable stuff from one's mouth, and to do so without opening one's
mouth widely (which might allow some undesirable substance to get in).
The link between disgust-type feelings and the kind of phonetic struc-
ture described above may seem so 'natural' and so self-explanatory that
it is difficult for a native speaker of English, Polish, Russian, German or
Yiddish not to be surprised when one finds out that, for example, in
Greek feu (phonetically nearly homophonous with the English phew of
316 Interjections across cultures
disgust) is an interjection of 'grief and anger' rather than disgust or
repugnance (cf. Kinchin-Smith - Melluish 1966:33).
Facts of this kind suggest that it would probably be crude and naive to
say, for example, that in interjections the sound [f] is linked, cross-
linguistically, with the meaning of 'disgust'. We might of course try to
cover ourselves by insisting that we are talking not about 'disgust' as
such but about 'disgust and similar feelings'. But are anger and grief
expressed in the Greek feu similar to those expressed in the Polish fu or
to those expressed in the Yiddish feh? If we do not spell out the meaning
in question more precisely the claim becomes meaningless.
One hypothesis which suggests itself is this: a physical gesture of oral
or nasal rejection may be linked, cross-linguistically, with the semantic
component 'I don't want this', or possibly, with a combination of
components: 'this is bad' and 'I don't want this'. It needs hardly to be
pointed out that the component 'I don't want this' may be as compatible
with 'anger' or 'grief' (that is to say, with feelings of the kind encoded
in the Greek feu) as it is with disgust, repugnance or revulsion.
It is clear, however, that the presence of this component does not
always lead to some interjection whose form indicates 'oral or nasal
rejection'; and when it does, usually additional components are
needed - different ones in different languages. Furthermore, it is
likely that not all languages have interjections whose meanings include
the components 'this is bad' or 'I don't want it'. But these are all, of
course, empirical questions, which must await further investigation.
To quote Darwin once more:
We have now seen that scorn, disdain, contempt, and disgust are expressed
in many different ways, by movements of the features, and by various
gestures; and that these are the same throughout the world. They all consist
of actions representing the rejection or exclusion of some real object which
we dislike or abhor, but which does not excite in us certain other strong
emotions, such as rage or terror; and through the force of habit and associa-
tion similar actions are performed, whenever any analogous sensation
arises in our minds. (Darwin 1955:261)
"The rejection of some real object which we dislike or abhor" can be said
to correspond to the postulated semantic components: 'I don't want it'
and 'this is bad'. The proviso that the stimulus in question doesn't
arouse in us "certain other strong emotions, such as rage or terror" seems
to point to a certain 'disinterestedness' of the feelings in question. There
is an important difference between thinking 'this is bad' and thinking
Emotive interjections 317
'this is bad for me'; concepts such as fear, frustration or anger imply
the latter, whereas concepts such as disgust or contempt imply the
former. One could perhaps venture to hypothesise that - "throughout
the world" - symbolic actions of oral and nasal rejection (including
interjections) are associated with the thoughts 'this is bad', not with the
thought 'this is bad for me'; but this hypothesis, too, requires empirical
verification.
3.2. 'General purpose' interjections
According to most dictionaries, many interjections encode no more than
an unspecified emotion, an emotion of any kind. For example, LDOTEL
describes the English oh as "an interjection used to express an emotion
(e.g. surprise, pain, or desire)", and Webster (1973) describes the English
ah as "an exclamation expressive of pain, surprise, pity, compassion,
complaint, contempt, dislike, joy, exultation, etc. according to the man-
ner of utterance". Other writers on the subject of interjections, too, often
insist that this or that interjection can express any feeling whatsoever,
depending on the context and intonation. For example, Rosten (1968:
274) lists no fewer than twenty-nine emotions which, he thinks, can
be expressed by the Yiddish interjection oy (I omit his examples):
(1) Simple surprise, (2) Startledness, (3) Small fear, (4) Minor sadness,
(5) Contentment, (6) Joy, (7) Euphoria, (8) Relief, reassurance, (9) Un-
certainty, (10) Apprehension, (11) Awe, (12) Astonishment, (13) Indig-
nation, (14) Irritation, (15) Irony, (16) Pain (moderate), (17) Pain
(serious), (18) Revulsion, (19) Anguish, (20) Dismay, (21) Despair,
(22) Regret, (23) Lamentation, (24) Shock, (25) Outrage, (26) Horror,
(27) Stupefaction, (28) Flabbergastation, (29) At-the-end-of-one' s-
wittedness.
But neither long lists of this kind nor short statements such as 'inter-
jection X expresses an emotion (any emo"tion)' account satisfactorily
for the fact that different interjections have different values. For
example, Rosten illustrates the different values of the Yiddish
interjections oy and ah as follows:
When you jump into cold water you cry oy! Then, enjoying it, say a-aah.
When you commit a sin, you revel in the pleasure, a-aah; then, realising
what you've done, you cry oy! (Rosten 1968:274)
318 Interjections across cultures
He concludes: "Oy, accordingly, can be used to express: ... " and he lists
the twenty-nine different emotions cited above. But, clearly, the illustra-
tion provided suggests something quite different from a list of twenty-
nine different emotions, including joy, euphoria and so on, and in fact
it can hardly be reconciled with such a list. I do not deny that oy can
be combined with all those different emotions; but surely, it must
also have a meaning of its own, a meaning different from the meaning
of ah and much more specific than simply 'any emotion whatsoever'.
I believe, however, that to capture that kind of meaning we must look
not for some suitable emotion term or terms but for an underlying
thought - a thought which would be consistent with the entire range
of emotions compatible with a given interjection. To show how this
can be done I will start with the Polish oj, homophonous with the Yiddish
oy and closely related to it, but much more restricted in use.
3.2.1. The Polish oj
OJ is one of the commonest Polish interjections. SIP describes it as "an
interjection strengthening the utterance, expressing pain, regret, fear,
admiration, resignation, confirmation, etc." This description suggests a
virtually unlimited range of use and no particular cognitive content. In
fact, however, the range of oj is limited, and its cognitive content is quite
specific, although difficult to spell out in words. The most common
collocations involving oj are probably these: oj, niedobrze 'oj, that's
bad', oj, boli! 'oj, it hurts!' and oj, nie 'oj, no'. These common colloca-
tions suggest clearly a negative bias. Intriguingly, however, oj can also
be used sometimes in a positive context; in particular, it can be used as
a 'moan' of admiration or delight.
Searching for a possible underlying thought which would be compat-
ible with all these different clues I am inclined to propose the following
tentative formula:
I feel something thinking about it
I feel like someone who thinks:
things can happen to people that people don't want
This combination of components presents the speaker as feeling rather
helpless and powerless. This is not necessarily a 'bad feeling', but it is
more likely to be linked to pain, regret, apprehension or resignation
than to positive feelings. When it is linked to positive feelings it
Emotive interjections 319
sounds as if the speaker was surprised at his good fortune. For example,
a sentence such as
OJ, jak tu
'OJ, how beautiful this place is!'
sounds a bit like a moan, a whine, or a squeak of delight.
In searching for the elusive invariant of this interjection we can
proceed as follows. First, we can note that it is often used to draw
attention to one's pain (as in the common complaint oj boli 'oj, it hurts');
second, that it is often used in other 'negative' contexts, such as a
situation of regret, apprehension or resignation; third, that it is never
associated with violent negative feelings, such as anger or despair; and
fourth, that when it expresses positive feelings, such as delight or
admiration, it has a somewhat squeaky or whiny character, as if a little
mouse or a little bird were expressing their feelings. One can't imagine,
for example, a lion having the kind of feelings which could express
themselves in an oj. The common theme seems to be something like
weakness or helplessness. For example, oj nie 'oj no' sounds like a
whine rather than like a protest, refusal or rejection. When the speaker
feels fully in control of the situation, oj can hardly be used:
a. OJ, niedobrze.
'OJ, that's bad.'
b. *OJ, swietnie.
,OJ, that's excellent.'
c. OJ, jak tu
'OJ, how beautiful this place is!'
Sentence (a) above is very natural because it expresses a negative evalu-
ation and suggests that the speaker is not in control of the situation;
sentence (b) suggests the opposite, and is infelicitous; sentence (c) ex-
presses a positive evaluation, but does not suggest that the speaker may
be in control, and it is felicitous. The speaker's delight conveyed in (c) is
coloured with surprise and it is quite compatible with an attitude expect-
ing something bad (rather than something good). Most characteristically,
however, oj is linked with helpless negative feelings, as in the following
examples:
OJ, nic dobrego z tej dziewczyny nie (Gomulicki) (SJP)
,OJ, this girl will come to no good.'
320 Interjections across cultures
OJ, ja w
Kocham a wzifl,c jej nie (Grudzinski) (SJP)
'OJ, how unhappy I am; where should I go?
I love a girl, and I cannot take her!'
Consider also the following examples from children's poems:
OJ, na niebo wyszly chmurki! (Szelburg-Zarembina 1970: 12)
'OJ, clouds have appeared in the sky!'
OJ, niedobrze - mysli swierszczyk -
To nie iarty z zimnym sniegiem! (Grodzienska 1970:30)
,OJ, that's bad - thinks the Cricket -
Cold snow - that's serious!'
OJ, to klopot wielki! Te male literki
Mylfl, Jasiowi jut: czy to A, czy to B
pr6tno pojfl,C chce, nie ani rusz. (Grodzienska 1970:84)
'OJ, that's a worry! those little letters are hard to tell apart: is it
an A, is it a B? Johnny can't understand it, he can't get it right.'
Kladzie mama koteczka do cieplego 16ieczka.
Bierze syna za OJ, masz kotku (Grodzienska
1970:81)
'Mum puts her kitten to bed-DIM. She takes him by the paw-
DIM. - OJ my kitten, you have a temperature!'
OJ, ratunku! slonko tonie! S10nko tonie! Co to
(Grodzienska 1970:10)
'OJ, help! The sun-DIM is sinking! What will happen?'
The negative bias of oj, which is not strong enough to prevent it from
co-occurring with positive feelings, is even stronger in the longer variant
of this interjection: ojej, (and in its childish derivate ojejku), which has
a somewhat exasperated tone. Another fairly clear difference between
ojej and oj lies in the more spontaneous character of ojej, which has to
be linked with a current (and, typically, sudden) thought. For example,
one could hardly replace oj with ojej in the sentence:
OJ, [*ojej) ja ... kocham a jej nie

'OJ, how unhappy I am! ... I love a girl and I cannot take her!'
Emotive interjections 321
which implies a long term reflection. By contrast, ole} is typically used
as a comment on what is happening at the moment, as in the following
examples:
O}e}, alei lecicie, ledwom was zlapal. - zawolal ostatkiem tchu.
(Brzoza) (SIP)
'Ole}, you're racing along. I scarcely caught up with you, he
panted.'
O}e}, tak ieby nie pomylic! - wzdychala Marysia.
(Swierszczyk) (SIP)
'Ole}, I'm so scared that I could make a mistake! - Marysia was
sighing. '
Both the similarity and the difference between ojej and o} are well
illustrated in the following passage, where o}ej is linked with a thunder-
storm and oj, with a shower. (Partly, it is no doubt a matter of the right
number of syllables; but it is also a matter of semantic plausibility.)
Idzie chmura - ciemna, bura, duia.
Oje}! Bedzie burza!
Szumi(!, gn(! skrzypi(! drzewa.
OJ! ulewa! (Szelburg-Zarembina 1970:22)
'A cloud is coming - dark, brown, big.
Ojej! There will be a thunderstorm!
Trees are rustling, bending, creaking.
OJ! There will be a shower!'
To account for both the similarities and the differences between ojej
and o} the following explication can be proposed for ole}.
o}ej
I now think: things can happen that I don't want
I feel something bad because of that
This formula differs from that assigned to oj by the element 'bad' in the
last component, by the direct reference to 'I' in the second component,
and by the added element of immediate thought: 'now' in the first
component.
322 Interjections across cultures
3.2.2. The Russian oj
It is interesting to note that a very similar interjection (also spelt oj)
exists in Russian, and that it has a semantic value very similar to that
of the Polish oj. For example:
OJ! Petr !vanovic, Petr !vanovic, nastupili na nogu! (Gogol')
(SSRLJa)
'OJ! Petr Ivanovic, Petr Ivanovic, you stepped on my toe!'
OJ, strasno! Ne govorite, pozalujsta, ne prodolzajte. (Ostrovskij)
(SSRLJa)
'OJ, that's scary! Please, don't speak, don't go on.'
Mamocka, oj, kak xoroso! (Gor'kij) (SSRLJa)
'Mummy, oj, that's excellent!' [lit. 'how good! ']
It appears, however, that the range of use of the Russian oj is wider, and
that its frequency is higher than that of the Polish oj. This is reflected,
for example, in the existence of the verb ojkat' (to say oj) in Russian.
Polish has no corresponding verb derived from oj, although it does have,
for example, the verb achac (to say ach) derived from the common
Polish interjection ach. (Russian, too, has a correspnding verb axat' /
axnut', as well as oxat' /oxnut': 'to say ax' and 'to say ox').
The wider range of use of the Russian oj than of the Polish oj can be
illustrated with the following example:
Potom, podojdja k divanu, ... s udivleniem skazala: - OJ, mja-
gko. (Simonov) (SSRLJa)
'Then, coming up to the couch, ... she said with surprise: - OJ,
soft. '
In Polish, one would not use oj in a similar sentence, although one could
exclaim, for example:
OJ, jak m ~ k k o
'OJ, how soft!'
?OJ, m ~ k k o (cf. OJ, niedobrze!)
'OJ, soft! (cf. OJ, (that's) bad!)'
The difference seems to be that in Polish positive evaluation can be
linked with oj only if it is sufficiently extreme to warrant something like
a squeak of delight. The Russian oj, on the other hand, can convey an
Emotive interjections 323
element of surprise, being linked with a sudden thought, ('I now think
X'), and can therefore be applied to relatively trivial situations. The
differences between the two ojs can perhaps be represented as follows:
oj (Polish)
1 think: things can happen to people that people don't want
1 feel something thinking about it
oj (Russian)
1 now think (know?) something
1 wouldn't have thought that 1 would think (know?) this
1 think: things can happen to people that people don't want
1 feel something because of that
The phrasing of the 'emotive' component is meant to reflect the greater
immediacy of the Russian oj, which sounds as if it was less controlled
than the Polish oj. This spontaneous, uncontrolled character of the
Russian oj seems particularly clearly reflected in the verb ojknut', 'to
utter oj'.
3.2.3. Ochs and achs
Och and ach, too, may appear to be all-purpose interjections in Polish.
SIP defines ach in very general terms: "an interjection expressing a
lively movement of emotion (lit. 'feeling')" and for och it suggests
virtually the same, although the phrasing is somewhat different: "an
interjection expressing a lively movement of emotion (wzruszenie, lit.
'being moved'), excitement, pain, impatience etc."
As one could expect, however, the two are in fact not identical, with
och sounding even more 'emotional' than ach, and being more likely
to be seen as an exaggerated outpouring of emotion. But here as
elsewhere, the apparent difference in strength is in fact due to an under-
lying difference in quality: there are some contexts when - regardless
of 'strength' - och is more felicitous than ach, or vice versa. For
example, och tak 'och yes' could not be used to check whether one has
understood the interlocutor correctly, whereas ach tak 'ach yes' could
very well be so used:
A: 'He is leaving.'
B: Ach tak? / *Och tak?
324 Interjections across cultures
In some contexts, both ach and och can be used, but with a clear
difference in meaning. For example, ach nie 'ach no' is likely to be
used to correct an error on the part of the addressee, whereas och nie
'och no' suggests, rather, that the speaker is rejecting some proposal. An
affirmative ach tak could be used when one hears something interesting
and unexpected, as one says in English is that so. But an affirmative och
tak cannot be used in that way.
A: 'He is leaving.'
B: Ach tak! / *Och tak!
On the other hand, an affirmative och tak! (but not ach tak) could be
used in reply to an eagerly awaited invitation or proposal:
A: 'Would you like to come?'
B: Och tak! / *Ach tak!
What these clues suggest is that ach has something to do with knowledge
(or with coming to know something), whereas och involves the speaker's
wanting rather than knowing. This difference between the two can be
represented as follows:
ach
I feel something (now)
because I know something
och
I feel something (now)
because I want something
These explications explain, I think, why one can use ach - but not
och - to ask for confirmation: it makes sense to ask the addressee to
confirm a fact but not to ask him to confirm that I want what I want.
However, the explications sketched above are lacking some compo-
nents which would account for the ironic use of nouns and verbs derived
from these interjections. This ironic use of ach and och can be illustrated
with the following examples:
wpadaj{l w achy i w ochy na widok ponczochy. (Tuwim) (SJP)
'they lapse into oohs and aahs (achs and ochs) at the sight of
stockings. '
Przybyla jakas niemiecka Korinna, pelna ach6w. (Chopin) (SJP)
'There arrived a German woman called Corinna emitting nothing
but aahs (achs).'
Emotive interjections 325
Wrocili nareszcie! i w kilku powiatach zakipialo jak w garnku na
odglos tej nowiny; a bliskie s(lsiedztwo ochalo i achalo z podium,
sluchaj(lc opowiadania pana Groszka 0 cudach ich rezydencji i 0
trybie ich zycia. (Plug) (SJP)
'At last they returned! And in several districts, the news evoked a
furious excitement; their neighbours on the podium oohed and
aahed, as they heard Mr Groszek's stories about the marvels of
their residence and their way of life. '
The point is that of all the Polish interjections, these two more than any
other are perceived as 'loud' and rhetorical, as if the speaker was trying
to publicise his or her emotions and to attract to them other people's
attention. For example, they seem much 'louder' (both phonetically and
semantically) than the English interjections ooh and aah, which would
normally be used in English translations of Polish sentences with och
and ach. This 'loudness' is no doubt due to the highly audible final
fricative consonant, which the English interjections in question don't
have. It should be noted that Polish has also purely vocalic interjections
o and a, which also correspond, roughly speaking, to the English oh and
ah, and which also differ in 'loudness' from och and ach. To account for
this 'loud' character of ach and och, I would postulate for them the
following additional component: 'I want someone to know about this'.
OJ is not perceived as 'loud' and an utterance opened with oj would be
much more likely to be represented in writing with a comma than with
an exclamation mark. It is much more common for ach than for oj to be
represented in writing with an exclamation mark, and even more com-
mon, for och, as in the following examples:
Och! zawolola z dobrze udanym zachwytem.
- Pani co dzien piekniejsza. (W. Kowalski) (SJP)
"'Och!", she exclaimed, with well-feigned delight.
"You are more beautiful every day.'"
Kamionkowa wydala okrzyk: och! i zemdlona padla na z i m i ~
(Dygasinski) (SJP)
'Mrs. Kamionek gave a cry: och! and fell in a faint on the
ground.'
Och! jak bolesnie, och! jak bolesnie,
ie dzien wczorajszy nigdy nie wskrzesnie! (Syrokomla) (SJP)
'Och! how painful, och! how painful,
That yesterday can never be restored!'
326 Interjections across cultures
It is worth noting that virtually homophonous interjections ox and ax
exist also in Russian, where they appear to be used with meanings very
similar to the Polish ach and och. A more detailed discussion of these
interjections, however, is beyond the scope of the present study.
4. Cognitive interjections
Cognitive interjections are global (unanalysable) expressions which ex-
press the speaker's mental state without reference to feeling or wanting.
Three examples from Polish, aha, oho and 0, will be discussed below
and will be compared with some related interjections from Russian
and English.
4.1. The Polish aha and Russian aga
SIP defines aha as "an interjection which expresses recalling, confirm-
ing, satisfaction, irony". In fact, aha by itself never signals satisfaction
or irony, but once the invariant of this interjection has been captured it is
easy to see why such attitudes can easily 'attach themselves' to it, in a
suitable context. 'Recalling' and 'confirming' are more directly relevant
to that invariant, but no more so than 'recognising' or 'understanding'.
What all of these different uses have in common is a kind of sudden
realisation, which can be captured in the following invariant: 'now I
know it'.
The following examples demonstrate, I think, the semantic correspon-
dence between the Polish aha and the expression 'now I know it'.
o czym mysmy mowili? ... Aha! Otoi chcialem was spytac ...
(Iackiewicz) (SIP)
'What were we talking about? Aha! [oh, that's right - now I
know it] I wanted to ask you '
Aha, jui wszystko rozumiem. (Prus) (SIP)
'Aha [now I know it], I understand everything now.'
Aha! ... a tuscie zlodzieje! (W. Boguslawski) (SIP)
'Aha! [now I know it] ... here you are, you thieves!'
Cognitive interjections 327
PowtarzaJ z przechwaJk{l. Aha! Nie na moje wyszlo? (Orkan)
(SJP)
'He was repeating boastfully: Aha! [now I know it] Wasn't I
right all along?'
I do not maintain, however, that the expression 'now I know it'
captures the full meaning of aha. As we will see below, Polish has
also another interjection (oho) to which the same (or a very similar)
component could be attributed and which nonetheless differs from aha
in its range of use. The main difference is due to the fact that in oho
the knowledge is due to perception, whereas in aha it is due either to
the understanding of something that the interlocutor has said, or to a
sudden realisation (that is, to a thought), or to a sudden discovery
confirming one's thoughts or suspicions. As a generalisation which
would cover all these different possibilities and which would nonethe-
less exclude cases where oho rather than aha would be said, I propose
the following:
aha (Polish)
I now know it
I have thought about it
Understanding, recall, realisation (through a thought process), confirma-
tion of one's suspicions - all these different situations where aha is
likely to be used, are compatible with this formula. Situations where
oho can be used but aha cannot, appear to be incompatible with it.
It is interesting to note that the corresponding Russian inter-
jection, aga, is used to indicate 'agreement or confirmation' (SRJa).
For example:
- My tebja na poruki voz'mem - skazal Fidel' - ja pogovorju s
muzikami.
- Aga, pogovori. (Dovlatov 1977: 135)
"'We'll vouch for you", said Fidel, "I'll talk with the blokes."
"Aga, (you) talk (then)." ,
- Mal' ciki, - sprosila Nadja, - vy nemnogo coknutye?
- Aga, - govorju, - my psixi. Kukareku. (Dovlatov 1977:136-
137)
"'Boys", asked Nadja, "are you both nuts?"
"Aga", I said, "we're psychos. Cock-a-doodle-do!" ,
328 Interjections across cultures
... ponjal?
Miska obradovalsja:
- Aga, ponjal! (Dragunskij 1968:207)
, " ... got it?"
Miska (answered) joyfully:
"Aga, I got it.'"
- fa, mama, sejcas byka s" est' mogu.
Ona ulybnulas'.
- Zivogo byka? - skazala ona.
- Aga, - skazal ja. (Dragunskij 1968:154)
'''Mum, I could now eat a bull."
She smiled.
"A live bull?" she said.
"Aga", I said.'
The Polish aha could normally not be used in corresponding Polish
sentences. As a first approximation, we could represent the meaning of
the Russian aga as follows:
I think: you want to know if I want to say the same (as you
now say)
you can know: I want to say the same
It is true that aga can also be used to show that one has come to under-
stand something, as the following example clearly shows:
- Xorosaja skola, - skazal on.
- Kakaja skola? - ne ponjal ja.
- Xudozestvennaja, - skazal on.
- Aga, - skazal ja. Xotja vse ravno ne ponjal.
(Dragunskij 1968: 83)
'''A good school", he said.
"What school?" I couldn't understand him.
"An art school", he said.
"Aga", I said, although I still couldn't understand.'
It appears, however, that this use of aga to show a sudden understanding
can be interpreted by the same semantic formula: presumably, the first
speaker hopes that the addressee understood him, and wants to know
if the addressee would say the same; and the addressee assures him
that this indeed is the case: 'you can know: I want to say the same (I
have understood).' Interestingly, the same (or homophonous) Russian
Cognitive interjections 329
interjection aga can also be used to express "a successful guess, joyful
surprise, derision, etc." (SRJa), as in the following example:
Teper' ty nasi Aga, drozi'. (Puskin) (SRJa)
'Now you belong to us! Aga, you're trembling.'
This seems very similar to the triumphant and derisive use of the Polish
aha, as in the Polish examples quoted earlier:
Aha! a tuscie zlodzieje!
'Aha! so here you are, you thieves!'
Powtarzal z przechwalkfl. Aha! Nie na moje wyszlo?
'He repeated with satisfaction: Aha! Wasn't I right all along?'
However, in the case of the Polish aha', the formula
proposed for the purely cognitive Polish aha seems to fit, too (perhaps
with an additional component 'I feel something because of it'). By con-
trast, in the case of the Russian 'triumphant aga(2)' the formula proposed
for the purely cognitive aga(l) does not fit, and despite the intriguing
semantic links between the two agas, perhaps we should agree with
the dictionaries of Russian (cf. SRJa and SSRLJa) which treat them as
different lexical items: aga
l
, (roughly, agreement/confirmation) and
aga
2
(roughly, satisfaction)
The formula 'I now know', which was proposed for the Polish aha,
might seem to fit the use of the Russian aga in sentences such as (a)
below, but not in sentences like (b):
a. Aga! Stiva! Oblonskij! Vot ion! - pocti vsegda s radostnoj
ulybkoj govorili, vstrecajas' s nim. (Tolstoj) (SSRLJa)
"'Aga! Stiva! Oblonsky! There he is!", they would say almost
invariably with a happy smile whenever they met him.'
b. Aga! - dognal tebja! postoj! - kricit naezdnik udaloj.
(Puskin) (SSRLJa)
"'Aga! I've caught you! One moment!" shouted the bold
horseman.'
The formula which would fit examples of both kinds (a and b) seems
to be this:
aga
2
now I can say it
I feel something good because of that
330 Interjections across cultures
The Polish aha implies that the speaker discovers some kind of missing
link, that is, comes to know something that a moment earlier he didn't
know; the Russian aga
2
implies that the speaker comes to be able to say
something, which a moment earlier he wasn't able to say. To capture the
link between aga
l
(the aga of agreement) and aga
2
we could now try to
rephrase the formula assigned earlier to aga
l
as follows:
aga
l
I think you would want to say:
'you and I say the same (about it)'
you can say it now
Interestingly, the formula assigned above to the Russian aga
2
seems to
fit also the English interjection aha, which, according to LDOTEL, is
"used to express surprise, triumph, derision, or amused discovery".
Triumph, derision, or amused discovery are all perfectly consistent
with the formula:
now I can say it
I feel something good because of that
It is true that LDOTEL mentions also surprise, which may have little to
do with this formula. As LDOTEL itself recognises, surprise can also be
expressed by other interjections, for example by gee and wow, whose
meaning is of course different from that of aha. It could be argued that
aha by itself does not signal surprise, although it may be compatible
with it; but perhaps the English aha should have a third component
referring to something like surprise:
aha (English)
now I can say it
I feel something good because of that
one wouldn't have thought I could say it
It should be noted that the English aha cannot be used as, roughly
speaking, a sign of affirmation ('yes'), as the Russian aga
l
can. In this
respect, the English aha is similar to the Polish aha. But the English aha
and the Polish aha differ from one another in another respect. The Polish
aha appears to be used much more widely and much more frequently
than the English aha. This difference in range and frequency is linked, I
think, with the fact that (as suggested by LDOTEL) the English aha
carries an emotional component (represented in my formula as 'I feel
Cognitive interjections 331
something good because of that'), whereas the Polish aha can be used
as a purely cognitive interjection:
now I know it
I have thought about it
An emotional component ('I feel something [good?] because of that')
can of course attach itself to this cognitive meaning and can be signalled
by the tone of voice, but it is not part of the semantic invariant of the
Polish aha.
4.2. The Polish oho
According to SIP, oho is "an interjection strengthening an utterance" and
either (a) "expressing admiration, approval, surprise" or (b) "empha-
sising a fact which has just been stated and which is usually evaluated
as negative".
In this section I will limit myself to the non-emotive (i.e., b) use of
oho. SIP is on the right track, I think, when it links oho with "a fact
which has just been stated". It is important to add, however, that in
this case the speaker notices the fact rather than realises (or recalls) it,
as in the case of aha. For example, one could say oho - but not aha -
on hearing remote thunder; one could also say oho in response to some-
thing the interlocutor says, but not to indicate comprehension (in which
case aha would be appropriate). Oho would indicate that the speaker
treats the interlocutor's utterance as an observational datum. For
example, it may convey the idea 'now I see that you are afraid of
your wife'. The main difference, then, is that aha links knowledge with
thinking, whereas oho links it with perception or observation. To avoid
using in the semantic formula the complex and language-specific term
'perceive' ('I perceive something') we can represent this link with
observation as follows: 'I think anyone who is here now can know it'.
aha
I now know it
I have thought about it
oho
I now know something
I think anyone who is here now can know it
332 Interjections across cultures
As these formulae suggest, there is also another, related difference
between oho and aha, which has to do with the contrast between 'old
information' and 'new information'. In the case of oho what is perceived
is 'new information'. In the case of aha the knowledge is new (in the
sense of being just acquired), but its content may have been 'old' (the
speaker may have suspected this content to be true before, or may have
known it and forgotten it). Hence the contrast between 'something' in the
explication of oho and 'it' in the explication of aha.
Furthermore, as pointed out by SIP, oho appears to have a negative
bias. This could be represented by means of the component 'I think this
is bad', but this might be too strong, as the negative evaluation is only
hinted at rather than clearly conveyed. A better phrasing of the relevant
component is therefore 'I think this can be bad'. This brings us to the
following formula
oho
now 1 know something (X)
1 think anyone who is here now can know it
I think this (X) can be bad
There is, however, one further respect in which this formula is perhaps
not entirely satisfactory, which is revealed by the comparison between
oho and the English expression: oh-oh (to be discussed shortly). The
point is that oh-oh (unlike aha) is something that the speaker might be
muttering to himself; by contrast, oho is deliberately 'loud', as if the
speaker was noting a 'public' fact and possibly trying to draw other
people's attention to it. This suggests the following further amendment
to the formula proposed:
oho
now I know something (X)
1 think anyone who is here now can know it
1 want someone to know it
1 think this (X) can be bad
The reader is invited to check the adequacy of this formula against the
following examples:
Przyjrza/a ~ uwazniej Paw/owi, kt6ry siedziaf osowialy i
milczg,cy. - Oho - zauwazyla - jestes dzis nie w humorze.
(Brandys) (SIP)
Cognitive interjections 333
'She looked more closely at Pawel, who sat silent and owlish.
"Oho", she remarked, "you're not in a good mood today." ,
Oho, juz przyszla zona za leb bierze. (Przynski) (SIP)
'Oho, your future wife's already got you by the throat.'
Oho, na mnie jui czas! (looking at a watch). Za minut
capstrzyk ... (Gruszecki) (SIP)
'Oho, I have to go. In a quarter of an hour we've got a demo ... '
Oho! znowu moraly! gderal dzien caly! (Rodoe) (SIP)
'Oho! More sermons! He'll be complaining all day!
It could be added that oho is clearly related to the particle/interjection 0
and to the emotive interjection hoho. For reasons of space I will restrict
myself here to looking briefly at the cognitive (non-emotive) use of o.
4.3. The Polish 0
The demonstrative (non-emotive) Polish 0 (quite different from the
emotive/vocative 0) can be regarded as a modal particle rather than an
interjection, because it rarely is used on its own. It can, however, be so
used (with a gesture).
As pointed out by SIP, "the interjection (particle) 0 can accompany
the gesture of pointing to something in space, and drawing special
attention to it". For example:
Tu szlo }akies wo}sko: 0, slady po czolgach. (Broniewska) (SIP)
'Some soldiers have been past here: 0, here are marks left by
tanks. '
Wilgot, zimno, 0, tu lezy koc, a pan nawet nie okry}e.
(Szaniawki) (SIP)
'It's damp and cold, 0, and here's a blanket and you won't even
wrap up.'
0, mucha leci do koni. (Galczynski) (SIP)
'0, there's a fly flying towards the horses.'
Niechie pan doktor bedzie laskaw powiesic futerko tuta} ... 0
tuta}. (Prus) (SIP)
'Doctor, would you like to hang your coat here ... 0 just here.'
334 Interjections across cultures
The following semantic formula can account for all these (and other
similar) examples:
o
I think anyone who is here now can know it
I want you to know it
(I do this [gesture] because of that)
Like oho, 0 is 'public' because it refers to a public fact, accessible to
any observer ('I think anyone who is here now can know it'). At the
same time, however, it is aimed at a specific addressee ('I want you to
know it'), whereas oho is directed at someone/anyone rather than at
'you' ('I want someone to know it'). Unlike oho, 0 does not imply any
evaluation (any 'I think this can be bad').
4.4. The English oh-oh
On the other hand, a negative component is clearly included in the
English interjection oh-oh (uh-oh), which is used when one perceives
that something bad and unforeseen is going to happen to someone (for
example, a runner is going to fall into a hole of which the speaker is
aware but the runner is not). The meaning of this interjection can be
represented as follows:
oh-oh
I now know something
I think something bad will happen now
I would not want someone to think that it will be very bad
As pointed out earlier, this is related to the Polish oho, but it is more
'private' as it lacks the components 'I think anyone who is here now
can know it' and 'I want someone to know it', and is more definitely
'bad', though not very bad ('something bad will happen' versus 'this
can be bad').
4.5. The Russian ogo
In the Russian translation of Franklin Folsom's children's book entitled
The language book, the meaning of the Russian interjection ogo is
explained with reference to an imaginary prehistoric scene:
Cognitive interjections 335
'Ux! - a prehistoric hunter might have exclaimed throwing down the deer
which he had dragged into his cave. And the members of his family started,
perhaps, to repeat this sound every time after some hard work. Ogo! the
hunter's daughter might have said when she saw what a big deer her father
had caught.' (Folsom 1963:25)
The meaning of the Russian interjections ux and ogo suggested in this
passage seems to be this:
ux
I can now say that I did something that I had to do
I felt something bad when I did it
I don't have to do it now
I feel something good because of that
ogo
I now know something
I think this is more than one would have thought
(I think this is good)
(I feel something because of that)
Ux does not belong to the domain under discussion and will not be
discussed here any further. As for ogo, the formula proposed above on
the basis of one imaginary example seems to fit quite well the examples
adduced in the dictionaries of Russian - except for the positive com-
ponent 'I think this is good', which apparently is not part of the invari-
ant. For example, it is compatible with the first example below, but not
with the second:
Peresmotrju narocno, skol' ko u menja deneg. Eto ot sud'i trista;
eto ot poclmejstera trista, fest'sot, sem'sot, vosem'sot ...
devjat'sot ... Ogo! za tysjacu perevalilo! (Gogol') (SRJa)
'I'll take a look how much money I've got. Here's three hundred
from the judge; and from the postmaster another three hundred,
six, seven, eight ... nine hundred ... Ogo! That's already more
than a thousand!'
Pal' ba byla ne dal'se, kak za verstu. Vse nastorozilis'. - Ogo! Ne
dremljut japoncy! - skazal Sancer, nervno ozivljajas'. (Veresaev)
(SSRLJa)
'There was shooting, not more than a verst away. Everyone be-
came alert. - Ogo! The Japanese are pretty active! - said Sancer,
stirring nervously.'
336 Interjections across cultures
If we also exclude from the hypothetical invariant the emotive compo-
nent 'I feel something because of that', we are left with two components:
I now know something
I think this is more than one would have thought
The examples of the Russian ogo adduced so far would also be com-
patible with the component of 'perception', which we have included in
the explication of the Polish oho ('I think anyone who is here now can
know it'). But ogo can be used much more widely than the Polish oho -
partly, no doubt, because it does not require a reference to current
perception. Consider, for example, the following examples:
Kostik sprosil:
- A razve byvajut xorosie bolezni?
- Ogo, skazal ja, - skol' ko xoces'. (Dragunskij 1968:28)
'Kostik asked:
"Can there be good illnesses?"
"Ogo", I said, "plenty." ,
The Polish oho could not be used in a similar context - partly, no doubt,
because the speaker is referring here to his knowledge but not to his
perception. In fact, examples of this kind suggest also that the knowledge
implied by ogo should perhaps be represented as 'I know' rather than
as 'I now know'. In the case of the Polish oho, the speaker's knowledge
has the form of a realisation: 'I now know'. The Russian ogo, too, is
often used in the situation of a sudden realisation, but apparently it
can also be used to express knowledge acquired earlier. This brings us
to the following semantic formula:
ogo
I know something
I think this is more than one would have imagined
Or, perhaps:
ogo
I can now say something
because I know something
I think this is more than one would have imagined
Conclusion 337
5. Conclusion
I have tried to show that interjections - like any other linguistic
elements - have their meaning, and that this meaning can be identified
and captured in rigorous semantic formulae. Formulae of this kind can
explain the range of use of a given interjection, and can account for the
differences in the use of different interjections. Are the interjections
themselves identifiable (by native speakers) on the basis of the proposed
formulae?
Emotive and cognitive interjections seem to differ in this respect
from the volitive ones. Volitive interjections such as shoo or psst in
English can be easily identified on the basis of the semantic formulae
assigned to them. On the other hand, emotive and cognitive explications
such as the Polish oj, och, aha, or oho, are harder to identify, and a
native speaker may be sometimes at a loss trying to link a semantic
formula to an interjection which it is supposed to explicate. In this
respect, interjections may be different from verbs or adjectives (and
perhaps more similar to modal particles). However, if formulae of the
kind proposed here are compared with traditional descriptions of
interjections (along the lines of "a word which can be used to express
different emotions depending on the context and on the intonation") their
higher explanatory value seems to be evident: they try, at least, to
account for the differences in the use and in the value of different inter-
jections, and they try to discover in each case an invariant which would
explain the unique range of use of every individual interjection.
The fact that the route back from the explication to the word may be
longer and more difficult to travel than it usually is in the case of major
lexical classes, should of course be noted, and its implications should
be explored. Perhaps we should conclude from it that different types
of linguistic signs have different psychological status. After all, an
interjection. is an equivalent of a full sentence. Perhaps a mental act
encoded globally in one phonologically tiny word is generally harder to
recognise (reconstruct) on a conscious level than an act encoded in a
more articulated linguistic expression? Perhaps a global sign such as an
interjection is in some sense more 'automatic' than a non-global sign,
such as a verb or an adjective?
We should also ask if the 'semi-automatic' character of (primary)
interjections is not related to their partial phonological motivation. It
338 Interjections across cultures
seems clear that in the domain of interjections certain types of sounds
tend to be linked, cross-linguistically, with certain types of meanings.
Correlations of this kind have often been suggested in the literature,
but they could not be seriously investigated before the question of the
semantics of interjections was placed, in a serious way, on the linguistic
agenda. Here as elsewhere the absence of a language-independent
semantic metalanguage has been a major stumbling block. It is easy
enough to say, casually, that if the German interjection pfui, the Danish
interjection fy or the Russian interjection fu all express disgust,
then sound symbolism is probably involved. But if one considers that
'disgust' itself is just an English word, without an exact equivalent in
German, Danish or Russian (see Wierzbicka 1986c; to appear, chap. 3),
it becomes clear that it is not a language-independent descriptive
category, which could be fruitfully used for a cross-linguistic study of
correlations between form and meaning.
On the other hand, operating with universal or near-universal semantic
chunks such as 'good' and 'bad', 'feel' and 'think', 'know' and 'hear'
we can at least start to investigate the meaning of interjections in a
cross-cultural perspective. Once we have begun to model their meanings
in a relatively culture-free semantic metalanguage, we can start to
investigate the extent to which sound symbolism plays a role in the
functioning of interjections. We can start to explore the universal and the
culture-specific themes in the semantics of interjections, and the inter-
play between the two. We can also start to explore, and to document,
different 'emotive styles' associated with different cultures and reflected
in language-specific systems of interjections. For example, it can
hardly be an accident that in a culture whose most prominent speech
acts include laments and impotent curses - the Jewish-Yiddish culture
(cf. Matisoff 1979) - the most prominent interjection appears to be
oy vay (oy veh), a linguistic 'symptom' of distress and helplessness,
whose meaning can be represented as follows:
oy vay
I now know that something bad happened
I would want to do something because of that
I can't do anything
I feel something bad because of that
It appears that languages differ in the kinds of emotions which they
have found worthy of encoding in special interjections. For example,
some languages appear to have special interjections in the domain of
Conclusion 339
fear, others in the domain of anger, and yet others in the domain of
sadness and distress.
In addition to qualitative differences of this kind, there are also
important quantitative differences. Firstly, some languages appear to
have much larger sets of primary interjections than others. Secondly,
the interjections used in different languages differ greatly in frequency.
This last point, though generally very difficult to document, can be
illustrated with reference to English and Russian. For example, in
Zasorina's (1977) megacorpus of one million running words of Russian,
the interjection t'fu was counted 19 times, and fu, 23 times, whereas in
Kucera - Francis' (1968) comparable corpus of American English,
interjections such as yuk, phew or pooh were not listed once. Compare
also the following data regarding the most common emotive interjections
of both languages:
Russian
oj 19
ox 83
ax 212
English
oh 119
ah 22
Striking as these numbers are, they don't show the full extent of the
difference, because Russian has one more common interjection, 0,
whose frequency cannot be assessed on the basis of the available data:
o is also a preposition in Russian, corresponding to the English about,
and to some uses of the English of In Zasorina's data, the word 0 has
the frequency of 4156, and we don't know how many of these belong to
the interjection.
Generally speaking, one would expect that in societies which discour-
age a spontaneous and uninhibited show of emotions, the use of primary
interjections would be more limited than in those where emotions are
shown freely and where expressive behaviour is valued rather than
discouraged. The contrast between English and Russian illustrated
above is a good case in point (cf. Wierzbicka, to appear). But of course
in this area as in others, much research is needed before any firm
generalisations can be reached.
Cross-cultural research in the area of emotion concepts lexicalised in
different languages has just begun (cf. for example Levy 1973; Lutz
1982, 1988; Solomon 1984; Rosaldo 1980; Gerber 1985; Wierzbicka,
to appear, chaps. 3, 4). Cross-cultural research in the area of emotion
symptoms lexicalised in different languages in the form of interjections,
is, one can hope, about to begin.
Chapter 9
Particles and illocutionary meanings
There are few aspects of any language which reflect the culture of a
given speech community better than its particles. Particles are very often
highly idiosyncratic: 'untranslatable' in the sense that no exact equiva-
lents can be found in other languages. They are ubiquitous, and their
frequency in ordinary speech is particularly high. Their meaning is
crucial to the interaction mediated by speech; they express the speaker's
attitude towards the addressee or towards the situation spoken about, his
assumptions, his intentions, his emotions. If learners of a language failed
to master the meaning of its particles, their communicative competence
would be drastically impaired.
The meanings embodied in particles are often remarkably complex.
Though these meanings can perhaps be expressed, in one way or another,
in any language, they are often so complex that if a particular language
does not provide any abbreviatory devices (like particles) the speakers
are effectively discouraged from ever expressing them. As Hymes
(1974a:1450-1451) put it, although "one could come to render anything
in any language, given sufficient time and trouble", nonetheless "it is
not the case that one can 'say anything in any language' if conditions
of acceptability and cost, as are always present in real situations, are
admitted". For this reason, particles, which provide generally accepted
ways of expressing complex pragmatic meanings at minimal cost, play
an essential role in co-determining the range of behavioural styles that
a given language makes available to its speakers.
The meaning of particles is often excruciatingly hard to state. Not
that very many assiduous attempts have been undertaken towards that
end; on the contrary, until very recently, the meaning of particles has
seldom attracted linguists' attention. Even in semantics, it would be
difficult to point to a more grossly neglected area. As John Locke said
three hundred years ago:
This part of grammar has been perhaps as much neglected as some others
over-diligently cultivated. It is easy for men to write, one after another, of
cases and genders, moods and tenses, gerunds and supines: in these and
the like there has been great diligence used; and particles themselves, in
342 Particles and illocutionary meanings
some languages, have been, with great show of exactness, ranked into their
several orders. But though prepositions and conjunctions, etc. [original
emphasis] are names well known in grammar, and the particles contained
under them carefully ranked into their distinct subdivisions; yet he who
would show the right use of particles, and what significancy and force
they have, must take a little more pains, enter into their own thoughts,
and observe nicely the several postures of his mind in discoursing.
(Locke 1690, 2:99)
Fortunately, the situation described by Locke is now beginning to
change, and particles have finally started to attract serious attention. In
particular, there has been a wave of serious publications devoted to
particles in German and Russian, both particle-rich languages (see, for
example, Kemme 1979; Weydt 1969; Weydt et al. 1983 for German;
Boguslavskij 1985; Nikolaeva 1985; Universite de Paris-7 1986 for
Russian; for a pioneering study of particles in a non-Indo-European
language, see Ameka, to appear). Nonetheless, I believe that Locke's
comments on the importance of particles, and on the need for their
investigation, are still worth recalling; as are his remarks on the method-
ology suitable for this purpose.
Neither is it enough, for the explaining of these words, to render them, as is
usual in dictionaries, by words of another tongue which come nearest to
their signification; for what is meant by them is commonly as hard to be
understood in one as another language. They are all marks of some action
or intimation of the mind; and therefore to understand them rightly, the
several views, postures, stands, turns, limitations, and exceptions, and
several other thoughts of the mind, for which we have either none or
very deficient names, are diligently to be studied. (Locke 1690, 2:99)
The role that Locke attributes to particles corresponds to what in
current terminology are known as 'illocutionary forces'. I have argued
on more than one occasion (see Wierzbicka 1972, 1976, 1980) that the
only possible way to represent these illocutionary forces accurately is
to decompose them: illocutionary forces are bundles of assumptions,
intentions and other more or less elementary 'postures' and 'turns' of the
mind. I think that there is a profound insight in Locke's assertion that
particles represent actions (actions of the mind), and that they are, there-
fore, abbreviations for whole sentences:
The instances I have given in this one [particle] may give occasion to
reflect on their use and force in language, and lead us into the contempla-
tion of several actions of our minds in discoursing, which it has found a
Particles and illocutionary meanings 343
way to intimate to others by these particles, some whereof constantly, and
others in certain constructions, have the sense of a whole sentence con-
tained in them. (Locke 1690,2:100)
If a particle condenses in itself an entire sentence, then the proper
way of stating its meaning must be to reconstruct this sentence. This
conclusion seems to follow automatically from Locke's discussion. Yet
Locke himself did not state it explicitly, and in his analysis of his
selected examples he was content simply to make some general com-
ments on the function of that particular particle. It was Leibniz who
first applied to particles the principle of substitutability (the crux of
Leibnizian semantics in general): if a particle contains in itself a con-
densed sentence, then to state the meaning of this particle one has to
reconstruct this sentence in extenso. The possibility (or otherwise) of
substituting the reconstructed sentence for the particle in question
provides an empirical test of the adequacy of the proposed explication.
For a proper explanation of the particles it is not sufficient to make an
abstract explication... ; but we must proceed to a paraphrase which may be
substituted in its place, as the definition may be put in the place of the
thing defined. When we have striven to seek and to determine these
suitable paraphrases [original emphasis], in all the particles so far as
they are susceptible of them, we shall have regulated their significations.
(Leibniz 1949:366-367)
The task of reconstructing one sentence which would correspond to
all the varied uses of a particle may seem unfeasible. In fact, however,
there is no reason to assume a priori that all the different uses of a
particle correspond to just one sentence (and just one mental 'posture').
Some particles may well have more than one meaning. But they will
not each have 'countless' different meanings. To quote Leibniz again:
Scholars have attempted to make special treatises upon the particles of the
Latin, Greek, and Hebrew; and Strauchius, a celebrated jurisconsult, has
published a book upon the use of particles in jurisprudence, where their
significance is of no small consequence. We ordinarily find, however, that
it is rather by means of examples and synonyms that they attempt to
explain them, than by distant notions. Further we can not always find a
general or formal signification for them ... which would satisfy all
the examples; but notwithstanding this we can always reduce all the uses
of a word to a definite number of significations. And this is what should
be done. (Leibniz 1949:365-366)
344 Particles and illocutionary meanings
The idea that the meaning of particles can be adequately elucidated
by means of examples, synonyms and translation equivalents, would
probably not be seriously defended by many linguists as a valid theoreti-
cal option, although the practice in question is still widespread. But the
notion that the meaning of particles can be adequately elucidated by
means of abstract formulae, totally unsubstitutable for the particles
themselves, continues to be put forward as a theoretical program, and
continues, I would add, to fail to produce empirically adequate clues to
the use of the particles discussed. (For illustrations and discussion, see
Wierzbicka 1986a.)
I think Leibniz was right: one can always reduce all the uses of a
particle to a definite number of significations (stated as paraphrases
capable of being substituted for the particles in various contexts); and
this is what should be done. On the other hand, the Leibnizian approach
to the study of particles has produced in the course of the last decade a
number of empirically based studies, which have revealed the meanings
encoded in a number of particles from a wide range of languages, and
which have achieved a degree of predictive power undreamt of in the
more traditional approaches to the study of particles. (Cf. in particular
Ameka 1986; Boguslawski 1986; Goddard 1979, 1986; Grochowski
1986; Harkins 1986; Wilkins 1986.)
The purpose of this chapter is to demonstrate this method of analysis
by elucidating the meanings of a variety of particles in English and
Polish. This will be done in the following order: English 'quantitative'
particles (first non-approximative ones: only, mostly and just, and then
approximative ones such as about, around, and almost); English 'tempo-
ral' particles (such as already, still, and yet); Polish 'temporal' particles
with similar - but different - meanings; and Polish 'quantitative'
particles (first non-approximative, then approximative ones), also with
meanings similar to, but different from, those of the English particles.
In all cases, paraphrases will be sought which would account correctly
_for the observable similarities, and differences, in the range of use.
English quantitative particles 345
1. English quantitative particles
English has a number of particles used for talking mainly, or exclusively,
about quantities. These particles rely crucially on components such as
'no more', 'no less', 'not much', 'it could be more' or 'it could be less'.
Some of these particles, and related expressions, purport to give precise
information, whereas others are used as estimates and approximations.
Compare, for example, the expressions as many as and more or less. The
meaning of the former can be represented as follows:
as many as, as much as
it is this
one could think: it is less
it is not less
The meaning of the latter would at first seem to be representable along
the following lines:
more or less
it could be this
it could be a little more than this
it could be a little less than this
It should be pointed out, however, that while as many (much) as is
really restricted to quantitative contexts, more or less is not. Thus, one
can say not only more or less 50 but also it is more or less the same.
This suggests that while etymologically, or perhaps metaphorically,
more or less is related to 'more' and 'less', synchronically and liter-
ally it might be better represented in terms of 'different' rather than
'more/less' :
more or less
it could be this
it could be a little different from this
it couldn't be very different from this
This example illustrates the tendency of quantitative particles to
develop metaphorical extensions, and the difficulty one encounters in
trying to separate quantitative senses from non-quantitative ones. In
what follows, I shall largely ignore this problem, trying to focus, above
all, on the similarities and differences between individual particles and
particle-like expressions.
346 Particles and illocutionary meanings
1.1. Non-approximative particles: only, merely and just
1.1.1. Only
As a first approximation, the meaning of only can be represented as
follows:
only
it is this (many/much)
it is no more
one could think: it is more
For example:
only fifteen, only a dollar
it is this (many/much): [15, $1]
not more
one could think: it would be more
In some contexts, the phrase 'not another' appears to correspond to only
better than 'not more', as in examples like:
Only Socrates runs. (cf. Goddard 1979)
Only Nescafe gives you that fresh roasted flavour.
and it seems justified to suggest a second meaning, along the following
lines:
Only Socrates runs.
this (person, thing, place ... )
not another
one could think: it would be another
It seems to me, however, that in cases of this kind the identification
is due to the definite noun phrase, whereas only itself implies, even
here, the component 'no more'. For example, consider the following
exchanges:
a. - Who's there?
-Me.
- You and who else?
- Only me.
English quantitative particles 347
b. - Who's there?
- It's only me.
In (a), only me implies 'no more people', whereas in (b) it appears to
be concerned exclusively with identification, not with the number of
people ('me, not somebody else'). At the same time, it should be noted
that even in (b), 'only me' implies a kind of limitation and (in Sapir's
terms) 'downward grading': 'it's me, not somebody else, nothing more
than me, you don't have to worry'. This reassuring ring of the phrase
'only me' can be accounted for if we posit for this use, too, the com-
ponent 'no more than this' (in addition to 'not another'), although in
this case more does not refer to a number and it may said to be meant
metaphorically. It is an open question whether such a metaphorical
interpretation of 'no more' should be regarded as a separate meaning;
it should be recognised, however, that a single explication of only in
terms of 'no other' would not be sufficient, and that all uses of only
appear to require something along the lines of 'no more'.
The same 'downward grading' quality of only is evident in its
adverbial uses, for example:
I was only joking.
(cf. ??/ was only speaking seriously.)
Clearly, the speaker doesn't mean here 'I was doing one thing, no
more', and the paraphrase 'I was joking, nothing else' seems more
apposite. But here, too, the phrase 'nothing else' (or 'not something
other than this ') by itself would not account for the downplaying
implication of the sentence (cf. 'it was no more than a joke').
My tentative conclusion is that while the uses of only considered here
can apparently be reduced to the formula 'no more; one could think: it is
more', this formula is open to two rather different interpretations. In
clearly quantitative contexts, for example in combination with numerals,
only has no 'down-playing' quality, because its component 'no more'
is taken in a literal sense. If, however, no quantifier is present (for
example, in combination with proper names), only can be interpreted
in two different ways: either as 'this person (thing, etc.); no more
persons' or as, roughly speaking, 'this person; no one (nothing) else;
nothing more'.
348 Particles and illocutionary meanings
1.1.2. Merely
Merely is closely related to only, but it appears to have two additional
components, which can be roughly described as 'this is not much' and
'this is not something important'. For example, the phrase 'merely 50
cents' implies, like only 50 cents, that the sum in question was 'no more
than 50 cents and one could expect it would be more'. In addition,
however, it dismisses the sum in question as something small and
unimportant, which only doesn't do. If only can be said to be 'restrictive'
or 'limiting', merely can be said to be also 'minimising' and 'deprecia-
tory' or 'dismissive'.
To begin with the 'minimising' aspect of merely, it can be represented
as 'not much', and thus contrasted with the 'limiting' character of only,
which we have represented in terms of 'not more'. Compare, for ex-
ample, the following sentences:
a. She gave him only two thirds of the sum he expected.
b. She gave him merely two thirds of the sum he expected.
Sentence (b) sounds odd because 'two thirds' doesn't sound like 'not
much'; but (a) is fine because only does not imply 'not much', only 'not
more' (in this case, not the whole sum).
Turning now to the 'dismissive' or 'depreciatory' aspect of merely we
can note that by calling someone's words 'mere words', 'mere promises'
or 'mere conjectures' one is not only 'minimising' their value ('that's
not much'), but also dismissing them as unimportant, not worth thinking
about. For example, making a suggestion or putting forward a hypothesis
one could well add, modestly:
This is only a suggestion.
This is only my hypothesis.
but it would be rather self-defeating to say:
This is merely a suggestion.
This is merely my hypothesis.
Similarly, the sentence She's merely a child could well be used in a
situation when one was trying to dismiss a girl's words as unimportant or
unreliable. In a situation when one was trying to protect a child from
excessive burdens one would be more likely to say She's only a child
(or She's just a child).
English quantitative particles 349
It is worth noting that merely, in contrast to only (and just), can
hardly combine with an imperative (whether explicit or implicit):
Bring just (only, *merely) two.
Just a moment! *Merely a moment!
The imperative implies 'I want X', but merely implies 'X is not worth
bothering about'; hence the semantic clash.
Finally, one can say, in reply to the question Who was there?:
Only John and me.
Only me.
Just John and me.
Just me.
but hardly:
?Merely John and me.
?Merely me.
On the other hand, the sentences It's only (/just) a game and It's merely
a game are both felicitous, and easily interpretable. Only a game implies
'no more than a game' and merely a game dismisses the game in ques-
tion as something unimportant.
I suggest, then, that merely can combine only with words or expres-
sions which can be interpreted as 'cultural synonyms' of 'unimportant',
'non-serious', 'non-real' or 'small', and as opposed to something 'impor-
tant', 'serious', 'real' or 'big'. For example:
This is merely a game (not serious business).
She's merely a child (not an adult).
Mere words/promises/conjectures (not deeds, not certainties).
Merely a fraction (not the whole thing).
And so on.
As a first approximation, we could consider, therefore, the following
semantic formula:
merely
no more than this
this is not much
one doesn't have to think much about it
350 Particles and illocutionary meanings
1.1.3. Just
According to Lee (1987), the English particle just has as many as four
different meanings: a 'depreciatory' meaning, a 'restrictive' meaning, a
'specificatory' meaning, and an 'emphatic' meaning. In my view, the
claim that just is polysemous is perhaps justified, but the number of
meanings posited is too high, and the labels (as well as accompanying
descriptions) do not show what these meanings really are, how they are
mutually related, or how they differ from those of related particles
such as merely or only. After all, one could equally well say that merely
is depreciatory, or that only is restrictive; but this wouldn't explain
how just, merely and only differ from one another, or why their ranges
of uses are different.
Let us start with two examples of the kind that Lee calls
,depreciatory' :
They're not serious - just a nuisance.
- Have you had any chest pain?
- Just a little bit. Not as much as I had before.
According to Lee (1987:378), in examples of this kind "the speaker
uses the particle to minimise the significance of some process. ... In
many cases in this corpus [of doctor-patient conversations] a particular
process is explicitly downplayed by comparison with some other
process". In my terms, the meaning in question can be formulated
as follows:
just
nothing else (= not something other than this)
this is not much
one could think: it is (would be) more
The label 'depreciatory' is misleading, because it suggests a negative
evaluative component, which is certainly not there (cf. for example Let's
have dinner together - just the two of us); and so is the label 'down-
playing', which seems to suggest something unimportant. But the term
'minimise' is roughly correct, and it corresponds in essence to the
proposed formula. For example, the component 'not much' assigned to
just (but not to only) explains the following contrast:
Only 47 people came.
?Just 47 people came.
English quantitative particles 351
Only 47 implies that more than 47 could be expected, but it doesn't
imply that 47 is 'not much'; just 47 does imply this and this implication
makes the sentence sound rather odd, because if the number of people
is small enough for them to be individually counted, then 47 would
appear to be a lot. Similarly:
She gave him only two thirds of the sum that he expected.
?She gave him just two thirds of the sum that he expected.
At the same time, we should note that just, unlike merely, has not
only a 'minimising' but also a 'precise' ring to it: just this implies not
only 'this is not much' but also 'nothing other than this', 'precisely this'.
It is partly for this reason, I think, that the phrase: Just the two of us
sounds intimate and cosy: it is not just a question of a small number,
but also of the identity of those included (there will be no outsiders,
nobody other than the two of us).
Generally speaking, if only is perfectly neutral, and merely is
depreciative, just easily lends itself to mildly positive (reassuring,
defensive, apologetic, even praising) interpretations. The reason is that
while 'small' can be easily viewed as unimportant, it can also be viewed
as desirable ('small is beautiful', 'small is safe', and so on). In the
absence of a depreciative component ('one doesn't have to think much
about this '), and in the presence of an identifying one ('no other than
this'), the minimising component 'not much' embodied in just lends
itself easily to positive interpretations.
Lee (1987:387) illustrates what he calls the 'restrictive' meaning with
sentences such as these:
You can get a B grade just for that answer.
[' II just get you to take your shirt off.
Just close your eyes.
It seems to me that sentences of this kind fit perfectly well the formula
proposed above, and that there is, therefore, no need to assign to the just
used in them a separate, 'restrictive' meaning. Lee himself is aware
that there are no firm grounds for distinguishing the two hypothetical
meanings, but he blames this on "the indeterminate nature of the
boundary between these two categories" (1987:387). A precise semantic
formula such as the one posited here allows us to dispense with the
alleged polysemy and reduce the 'depreciatory' and 'restrictive' senses
of just to one. For example:
352 Particles and illocutionary meanings
I' II just get you to take your shirt off.
I'll get you to take your shirt off
nothing else
this is not much
one could think: it would be more
The 'specificatory' sense of just is illustrated with sentences like
these:
It gets itchy just under the eyebrows.
He'd just come out of hospital.
According to Lee (1987:389), in sentences of this kind a locative or
temporal expression refers to an area which is sharply defined at one
end but undefined at the other, and "the function of just ... appears to be
to identify only that marginal phase ... close to the sharply specified
[part of the] area". It seems to me that this is an insightful observation.
The ideas of 'not much' and 'not some other' appear to be relevant here,
too, but their interpretation is somewhat different. For example, just
under the eyebrows implies 'not much under the eyebrows', so much so
that the place in question can be thought of as the same (no other) place
as where the eyebrows are. Similarly, just after he'd come out of hospital
implies 'very little (not much) after' - so little that the time in question
can be thought of as the same as (not other than) when he came out of
hospital. Since places and times can be thought of as areas and periods,
as well as 'points', two different points in either space or time can be
thought of as 'the same place' or 'the same time'. For example:
just under the eyebrows
where the eyebrows are
not in another place
it is under (the eyebrows)
not much (not much under)
one could think: it would be more
just after he'd come out of hospital
when he came out of hospital
not at some other time
it was after (he'd come out of hospital)
not much (not much after)
one could think: it would be more
English quantitative particles 353
In a sense, then, this use of just does indeed 'specify' a certain place
or time - but not in the sense in which the particle right (as in it was
right here) specifies a certain place. The 'specificatory' just 'specifies' a
place or time by ignoring (downplaying, minimising) the distance
between two points in space or time, and this is how it is related to
the 'depreciatory/restrictive' just.
The so-called 'emphatic' just occurs in sentences such as the follow-
Ing ones:
It was just incredible.
I just can't keep going.
I just can't understand it.
Lee (1987:394) comments: "Here we have moved a long way from the
depreciatory just. In fact, the particle appears to have precisely the
opposite effect - to emphasise the expression with which it enters
into construction". And yet, Lee feels that there are close links between
these different and apparently opposed senses of just. It seems to me that
precise semantic formulae phrased in the natural semantic metalanguage
can show what these links are. In the two sentences below:
I just don't like it
I just can't understand it.
it is clearly counterintuitive to say that just is used in two very differ-
ent, indeed opposed, meanings. Intuitively, the link between these two
alleged senses seems very clear, and it can be stated as, roughly:
this is all I can say about it (e.g. I don't like/can't understand it)
I can't say anything else
I know: this is not much
one could expect more
But if we put it like this, it transpires that, in essence at least, the same
semantic formula which was proposed earlier for the so-called 'deprecia-
tory', 'restrictive' and 'specificatory' senses of just applies also, mutatis
mutandis, to the so-called 'emphatic' just:
It is just incredible.
It is just impossible.
I just can't understand it.
354 Particles and illocutionary meanings
just
1 can't say anything other than this
(e.g. it's incredible/impossible, I can't understand it)
(I know: ) this is not much
one could think: it would be more
The 'emphatic' sense of just appears to occur always in combination
with the idea of impossibility: I just can't, it is just impossible, it is just
unbelievable (i.e. one just can't believe it), it is just gorgeous (i.e. 'I
can't say anything other than this', 'I can't say another word'), and so
on. The phrase (I) just can't ... implies, generally, that 'I can't say
anything other than this; this is not much, but 1 can't'. The implication
is that the speaker is overwhelmed and lost for words; hence the impres-
sion that the message is, in the speaker's eyes, 'overwhelming', and that
the speaker is 'emphasising' it, not 'downplaying' it. But it seems
that this effect can be accounted for in terms of essentially the same
semantic formula which has been assigned to the other uses of just.
1.2. English approximative particles
There is no sharp line separating 'quantitative' particles from 'approxi-
mative' ones, because a particle, or particle-like expression, can be
'quantitative' and 'approximative' at the same time. This applies, in
particular, to the particle-like expressions at least, at the most, no less,
and no more, as used in the following sentences:
At least fifty people were there.
The dining-room seats forty at (the) most.
No less than fifty chairs were crammed into the tiny room.
No more than fifty people came to the plenary session.
at least
it is not less than this
it could be more
one could think: it would be less
at (the) most
it is not more than this
it could be less
one could think: it would be more
English quantitative particles 355
no less
it is not less than this
it could be more
one could think: it would be less
no more
it is no more than this
it could be less
one could think: it would be more
Like only, all of these expressions refer not only to quantifiers but also
to expectations, and they imply that the actual quantifiers are different
from the expected ones. At the same time, however, they imply that the
speaker is making an estimate, whereas only implies precise information.
It is important to stress that no less and no more act as approximatives
only when they are not followed by any further quantification. It is by no
means self-contradictory to say, for example, no less and no more than
twelve. But this does not mean that no less on its own does not imply
'it could be more', or that no more on its own does not imply 'it could
be less'.
In this section, I shall consider a number of 'approximative' words
and expressions, including around, about, approximately, roughly,
almost, and nearly, trying to reveal the similarities and the differences
between them.
1.2.1. Around and about
Around 50 people came. / About 50 people came.
Come around five 0' clock. / Come about five 0' clock.
Dictionaries usually 'define' around via about (and/or via approxi-
mately), and vice versa. And indeed, in many contexts, the two words
seem interchangeable. A rough gloss for all three is suggested by another
'approximative' expression, which can often - though by no means
always - be substituted for them: more or less.
Fifty people came, more or less.
(? )Come at five 0' clock, more or less.
But do around and about really mean the same? And what do they mean?
356 Particles and illocutionary meanings
Sadock (1981), having discussed about In some detail, formulates
his main conclusions as follows:
I suggest that about be given the following definition: A sentence of the
form about P is true just in case P is a quantitative proposition and there
is a possible world not very different from the real world in which P is
true. (Sadock 1981:267)
But this definition cannot be regarded as correct, if only for the follow-
ing reason: about is not restricted to 'quantitative propositions', and in
fact it differs in this respect from around. Consider, for example, the
following dialogue:
- Is the food there any better than in our college?
- About the same. / *Around the same.
I propose that this difference between the purely 'quantitative' around
and the 'quantitative' or 'qualitative' about can be accounted for by
defining around in terms of 'more' and 'less', and about, in terms of
'different'. As a first approximation, I would propose this:
around
it could be this
it could be a little more
it could be a little less
about
it could be this
it could be something a little different from this
Sadock (1981:262) has also claimed that all English 'approximatives'
have the conversational implicature 'not P' , i.e. 'not exactly P'. It seems
to me, however, that this is simply not true. By saying around twenty or
about twenty the speaker does not wish to imply that the actual figure
was different from twenty. He implies that the actual figure may be a
little different from twenty, but not that it has to be different. This is
why in the proposed explications of around and about I have included
the component 'it could be this'.
But there are also other differences between around and about. One of
them consists in the fact that around, but not about, is able to apply to a
whole area, or period, surrounding a point in space or time, as in the
following sentence:
Hats of this kind were worn in Paris around 1880.
English quantitative particles 357
If we replaced around with about, the sentence would imply that
what is meant is one particular year, not a few years, as in the case of
around 1880. In fact, about can hardly apply to a period at all; rather,
it is used as a guess, an estimate, a 'shot' at one particular point in time:
Hats of this kind first appeared in Paris (in) about 1880.
To account for this difference between about and around, we could
differentiate the relevant components as follows:
around
it could be this
it could be a little more than this
it could be a little less than this
about
it could be this
if not this, it could be something a little different from this
This means that about applies to one point: 'this', or to some other point
in the area nearby; but around can, in principle, apply to a number of
points in the area around the reference point. Of course such an extended
interpretation can be excluded by a particular context, such as around
50 people or about 50 people, where the difference in question gets
neutralised, but given an appropriate context the extended interpretation
is always available for around, and not for about. For example:
The next few days were very hot, so we returned to our pattern
of resting during the hottest part of the day and around
[/?about] midnight ... (Facey 1981:144)
Another difference between around and about has to do with the idea
of 'rounding' encoded in the former, but not in the latter. While one
would be unlikely to say either ?around 87 people or ?about 87 people,
it is easier to say about 27 or 28 people than around 27 or 28 people,
and easier still to say about 6 or 7 people than around 6 or 7 people.
In fact, around 6 or 7 people sounds as if one were rounding some
fractional number, such as six and a half, or six and three quarters. To
account for this difference between around and about, I would postulate
the following component for around:
I say X, not something a little different from X
because it is easy to think of X
358 Particles and illocutionary meanings
No such component would be postulated for about. The fact that one is
unlikely to say about 87 people can be explained without positing any
such component: it is simply a bizarre thing to do to mention such a high
number as a rough estimate; if the figure was based on counting then
there would be no need for about, and if it was not based on counting
then the estimate would probably be made in tens rather than in single
units (e.g. about eighty or ninety people). But when one says about 6
or 7 people one is trying to be accurate as far as possible, and no process
of 'rounding' is implied. In sum, the proposed tentative formulae for
around and about read as follows:
around
it could be this
it could be a little more than this
it could be a little less than this
it couldn't be much more than this
it couldn't be much less than this
I say this (number), not another (number),
because it is easy to think of this (number)
about
it could be this
if not this, it could be something a little different from this
it couldn't be something very different from this
1.2.2. Approximately
Approximately is similar to around in implying a process of 'rounding'.
One would be unlikely to say:
?Approximately 7 people came.
?Approximately 17 or 18 people came.
A feature which separates approximately from both around and about is
its greater degree of abstraction and conceptual complexity (reflected
in its stylistic 'bookishness' and formality). For example, approximately
can apply to relationships between sizes or dimensions rather than to
straight numbers:
This line is approximately (/*around) twice as long as that.
This block is approximately (/*around) three times as long as it
is wide.
English quantitative particles 359
The action was repeated five times, at approximately (I*around)
equal intervals.
Being a more 'learned' and abstract word, approximately suggests
also a degree of control and mental discipline. It doesn't sound like a
rough estimate, or a rough 'shot', which could reflect a lack of concern
for accuracy, but rather, like a purposeful device, revealing a respect for
precision even at times when the speaker feels precision is not called
for, or not possible.
Needless to say, it is not easy to reflect such subtleties in semantic
formulae. As a first approximation, the following formula may be
considered:
approximately
if it is more than this, it is not much more than this
if it is less than this, it is not much less than this
I say this word, not another word, because it is easy to think of
this word
I don't want to say 'it is this'
because I know it is something a little different from this
It is interesting here to compare approximately with exactly, which
may seem to be a kind of 'anti-approximative', symmetrically opposite
in meaning to approximately. In fact, however, there is an important
difference between the two (in addition to the obvious similarities):
exactly, like about, can appear in purely qualitative contexts:
It is exactly the same.
whereas approximately cannot:
?It is approximately the same.
To account for this fact, we can assign to exactly the following formula:
exactly
one can say this
I know: someone can think:
'it is not this, it is something a little different from this'
I don't want anyone to think this
one can say this, not something a little different from this
360 Particles and illocutionary meanings
1.2.3. Roughly
Dictionaries (for example both Webster 1977 and OED 1933) equate
approximately and roughly with one another. For example, Webster
(1977) says: "rough - approximate (a rough guess)". However, in the
very example provided to illustrate this presumed identity, approximate
couldn't be substituted for rough (*an approximate guess). Similarly,
one ~ have a rough estimate but not an *approximate estimate.
Generally speaking, roughly implies a quick action, and a willingness
to sacrifice precision in favour of convenience, ease and simplicity.
Approximate has no such connotations, and on the contrary, while sig-
nalling a lack of accuracy it manages to convey a respect for accuracy.
Another important difference between roughly on the one hand and
approximately (and, for that matter, around) on the other, has to do with
the qualitative, non-quantitative character of roughly. The sentence:
This block is roughly twice as long as it is wide.
sounds fine, but the following two sound a little less felicitous:
It was roughly five 0' clock.
There were roughly 50 people present.
Furthermore, roughly can apply to purely qualitative matters, as in the
following sentence:
My idea, roughly, is this.
Approximately is not restricted to numbers and quantities, but it does
seem to be restricted to matters of form, where an accurate reproduction
of all the parts is possible. For example, one can say, I think:
The meaning of the word X can be stated, approximately, as
follows.
if one assumes that there is a unique optimal semantic formula which
one may try to approximate, better or worse. One can hardly say,
however:
?My idea is, approximately, this.
Approximate(ly) implies that accuracy is possible, rough(ly) doesn't. One
can do a rough sketch for a future painting, but not an approximate
sketch. Finally, roughly is much further from the target than approxi-
mately: while approximately implies that what is said is no more than a
English quantitative particles 361
little different from what is true, roughly implies a difference that can be,
and probably is, quite considerable. For example, a rough sketch is in
some respects like the real thing, but it is in all probability more than a
little different from it.
roughly
it is not this
it is something like this
I say this because it is easy to say this
I want to say something that one can say quickly
1.2.4. Almost and nearly
Almost seems to have two rather different uses which can be called, very
roughly, 'gradual' and 'non-gradual'. These two uses can be illustrated
with sentences like these two:
He is almost bald/blind.
He almost killed her.
(gradual)
(non-gradual)
Almost} implies, rather like about, that one is describing a certain state
or situation in a way which is slightly inaccurate. Almost
2
occurs in
sentences referring to various 'narrow escapes'. In the former case,
omitting almost would make the sentence a little different from what
is true, while in the latter, it would make it patently false. For example,
the difference between 'being bald' and 'being almost bald' is slight,
whereas the difference between 'dying' and 'almost dying' is very
considerable. The labels 'gradable' (or 'gradual') and 'non-gradable', (or
'non-gradual '), however, are not particularly apposite, since the former
applies also to numerical contexts, as in almost twenty, where strictly
speaking there is nothing 'gradable' or 'gradual'. But it is not labels
that are important but semantic formulae. If we assume that there are
two distinct (though of course related) senses of almost we could
propose for them the following two explications:
almost}
(a) one can't say this
(b) if I said this
it would be no more than a little different from what one
can say
362 Particles and illocutionary meanings
almost
2
(a) one can't say: this happened
(b) if something was no more than a little different from what it
was one could say this
It is worth noting here that in some languages, for example Polish
(see section 4.2 below), there are two separate lexical items correspond-
ing to these two different uses of almost:
On jest prawie lysy.
'He is almost bald.'
On 0 malo jej nie zabil.
'He almost killed her.' (lit. 'By a little he didn't kill her. ')
Since, however, English doesn't distinguish lexically the two senses
corresponding to prawie and 0 malo nie, good methodology requires that
we should at least make an effort to find a unitary semantic formula for
the two uses, or for the common core of these two uses. The following
rough formula suggests itself as a possibility:
if something was no more than a little different from what it
is (was)
one could say this
In support of such a unified formula, one could point out that while there
is more than a little difference between being killed and being 'almost
killed', there may well be no more than a little difference between a
shot that kills and a shot that 'almost kills' (but misses); or between a
rage that leads to a killing and a rage that almost leads to a killing.
However, assuming that we have reduced the two apparently different
senses of almost to one, we still have one very serious problem to re-
solve: we have to explain why almost - in contrast to around, about and
approximately - seems to approximate the target, so to speak, from
below, not from both sides. An expression such as around 50 can refer
to situations where the actual figure was more than 50 (say, 52 or 53),
whereas almost 50 cannot be so used. This difference between almost
and around could of course be easily accounted for if we posited for
almost one additional component:
it is a little less than this
But this simple solution runs into difficulties when one considers
sentences such as:
English quantitative particles 363
Almost nothing was left.
Almost nobody came.
Even if we did posit two different meanings for almost, it would not
help us, since in these sentences almost is not used in the sense of
narrow escape; clearly, it is not almost
z
but almost1.
Sadock (1981) has argued against the inclusion of the component 'less
than X' in the semantic representation of almost, and I think that the
acceptability of expressions such as almost nobody and almost nothing
supports his position in this respect. On the other hand, his conclusion
that almost means virtually the same thing as about and around (roughly,
'not very different from what it is ') can hardly be accepted, in view of
the radically different interpretations of expressions such as almost 50
(definitely no more than 50 and in fact less than 50) and around 50
(could be a little more than 50).
Sadock tried to account for these differences in interpretations in terms
of the allegedly quantitative nature of about. We have seen, however,
that about is not restricted to quantitative contexts any more than
almost is. At the same time, around, which is restricted to quantitative
contexts, is similar to about, in implying a range on both sides of the
target point.
around 20
about 20
could be 20
could be a little less
could be a little more
could be 20
could be a little less
could be a little more
almost 20 a little less than 20
couldn't be 20
couldn't be a little more
I believe that the clue to the dilemma is provided in Paduceva's
(1985:74) observation that almost (or rather, its Russian near-counterpart
pocti) includes a negative component as well as a positive one. A
sentence such as
Almost 50 people came to the party.
conveys the idea that one couldn't say (truly) of any 50 people that
'these people came to the party'. The sentence:
364 Particles and illocutionary meanings
Around (about) 50 people came to the party.
doesn't imply that. I suggest that it is this negative component of almost
which creates the impression that in numerical contexts almost implies
'less than'. Around and about do not have such a component, and this is
why they never appear to imply 'less than'.
As a final point concerning almost, it should be mentioned that like
around (and unlike about), it implies a process of 'rounding'. One can
hardly say: ?Almost seven people came although one can very well
say: Almost twenty people came. Almost requires, therefore, the 'round-
ing' component:
I say this because it is easy to think of this
Sadock (1981 :267) asserts that "with the exception of a few idioms
such as about to and just about, almost can occur wherever about can".
But in fact one can hardly substitute almost for about in a sentence
such as:
About six or seven people came.
The 'rounding' component postulated here for almost accounts for this
difference. Sadock notes also that if, say, the actual number of demon-
strators on a particular occasion was 950, then the sentence:
Almost 1000 demonstrators picketed.
seems truer than:
Almost 990 demonstrators picketed.
The 'rounding' component posited here explains such facts. '1000' may
well seem the closest number bigger than 950 that could be chosen on
the grounds of being 'easy to think of' (in comparison with '950'). But
there is no reason why '990' should be thought of in such terms.
However, unlike around, almost is not in any way restricted to
numbers, and unlike approximately, it is not restricted to formal
relationships. One can say:
This block is almost twice as long as it is wide.
but one can also say:
At such moments, she seemed almost pretty.
English quantitative particles 365
whereas one can't say:
At such moments, she seemed *approximately (*around, *about)
pretty.
Interestingly, almost differs also in this respect from nearly, which is
otherwise analogous to almost in many respects:
*At such moments, she seemed nearly pretty.
The fact that one can say:
We are nearly (*around, *about, *approximately) there.
suggests that nearly is not restricted to numbers or numerical relation-
ships. I presume that the reason why one can't say nearly pretty is that
nearly is so to speak 'upward graded' (cf. Sapir 1949), in other words,
it is viewed in terms of 'adding more of the same'. But it is not clear
what exactly would have to be added to make someone pretty if they
are 'almost pretty'.
To put it differently, nearly appears to refer to a process, whereas
almost can be purely stative. For example, one would be more likely
to say (a) than (b):
a. When I first saw her she was almost naked.
b. When I first saw her she was nearly naked.
The phrase 'nearly naked' is not impossible but it implies that one
watches the process of undressing, as in:
By that time, she (the stripper) was nearly naked.
To account for this processual character of nearly we could posit for it
the component:
if a little more of the same happened
one could say this
But nearly is also 'upward graded' in a different sense, which can be
illustrated with the following contrasts:
nearly everything / *nearly nothing
nearly everyone / *nearly no one
almost everyone / almost no one
almost everything / almost nothing
366 Particles and illocutionary meanings
To account for these contrasts we can (as a first approximation)
postulate for nearly the component 'it is a little less than this', which,
as we saw, cannot be assigned to almost.
Thus, the following two semantic formulae can be proposed for almost
and nearly:
almost
one can't say: it is this
if something was no more than a little different from what it is
one could say this
I say this like this
because it is easy to say it like this
nearly
one can't say: it is this
if no more than a little more of the same happened
one could say this
it is a little less than this
The analysis of almost proposed here is very different from the
'radical pragmatic' one proposed in Sadock (1981). According to
Sadock, "the meaning of almost is such as to make a statement of the
form almost P true just in case there is a possible world in which P is
true that is not very different from the real world" (1981:258-259).
Sadock believes that all the aspects of the use of almost can be explained
in terms of this meagre definition strengthened only by 'heavy doses
of Gricean pragmatics'. But Sadock's definition would apply to nearly
as much as to almost. It would also apply to about and to many other
'approximatives' in English and in other languages. So how can such
a definition, even combined with the heaviest possible doses of
Gricean pragmatics, account for the differences in the range of use of
all such expressions?
I contend that what is needed to account for all such differences,
as well as similarities, in the use of related expressions, is a 'radically
semantic' approach, not a 'radically pragmatic' one.
English temporal particles 367
2. English temporal particles
English has a number of temporal particles, related to one another in
rather intricate, and certainly intriguing ways. There is a considerable
literature devoted to these particles or to their counterparts in other
languages, in particular, in German, Polish, and Russian (cf. e.g.
Bankowski 1971, 1975a,b, 1976, 1977; Doherty 1973; Grochowski 1986;
Konig 1977; Moiseev 1978; Pasicki 1976; Shetter 1966; Traugott -
Waterhouse 1969). I do not survey this literature here, partly for reasons
of space, and partly because, with the exception of Grochowski (1986),
it doesn't propose any explicatory semantic formulae which would
attempt to account for the range of use of the particles under discussion.
Two of these particles, already and still, appear to be mutually
symmetrical, at least in the sense that they both refer to an expected
change, and relate this change to an expected time: in the case of al-
ready, the change occurred, and it occurred before the expected time,
whereas in the case of still, the change did not occur before the expected
time, and in fact, it did not occur at all.
He still hasn't come.
it (the situation) is the same (as before this time)
one could think:
something would happen
it would not be the same because of this
it would happen before this time
it did not happen before this time
He has already come.
it (the situation) is not the same (as before this time)
one could think:
something would happen
it would not be the same because of this
it would happen after this time
it happened before this time
In the case of sentences with stative predicates the reference to
an event is implicit, but it is still there; for example (cf. Traugott -
Waterhouse 1969:302):
He is still young (*old).
he is the same (as before this time)
368 Particles and illocutionary meanings
one could think:
something would happen
he would not be the same because of this
it would happen before this time
it didn't happen before this time
He is already old (*young).
he is not the same (as before this time)
one could think:
something would happen
he would not be the same because of this
it would happen after this time
it happened before this time
The word 'it' in the last two lines of these explications refers to some
expected change. Consequently, these explications explain why the
starred variants are not felicitous: it makes sense to refer to an
expected change from 'not old' to 'old' or from 'young' to 'not young'
but not the other way around.
Still is also closely related to yet, as they both refer to an expected
change which hasn't happened, or possibly hasn't happened. But in the
case of still, the change was expected to happen before a certain time,
whereas in the case of yet it was expected to happen at some unspecified
time, not necessarily before the reference time. For example:
It still hasn't happened.
it (the situation) is the same
one could think:
something would happen
it would not be the same because of this
it would happen before this time
it didn't happen before this time
It hasn't happened yet.
it didn't happen before this time
one could think: it would happen
one can think: it will happen
It is easier to interpret the sentence with still than the sentence with yet
as impatient; and it is easier to interpret the sentence with yet than the
sentence with still as confident: if the change was expected to occur
before now, and hasn't, the speaker may well feel impatient (and perhaps
English temporal particles 369
even doubtful as to whether the change will occur at all); if the change
was expected to occur at some time, not necessarily before now, there is
no reason to feel impatient, and there is every reason to be confident
that it will occur later.
Another interesting difference between yet and still has to do with the
fact that yet occurs, as a rule, with sentences which are either negative
or interrogative, whereas still is also used in affirmative declarative
sentences. This difference is related to the fact that still indicates the
continuation of a state ('it is the same as before') whereas yet does not
have this component:
He still hasn't come.
it (the situation) is the same
it didn't happen
it didn't happen before this time
He hasn't come yet.
it didn't happen
it didn't happen before this time
This explains the following contrast in acceptability:
Has he come yet?
*Has he still come?
The starred sentence is, semantically, self-contradictory, because the
speaker appears to be asking two incompatible questions at the same
time:
a. is the situation the same (as before now)?
b. did he come (before now)?
But 'his' coming would mean that the situation is not the same. Yet does
not refer to any sameness of the situation, and consequently, the unstar-
red sentence above is not self-contradictory; it implies only (b), not (a).
One feature which all three particles discussed here share is their
relative character: none of them allows for the event to be 'dated', they
only relate the (unspecified) time of the event to the reference time.
Consequently, they don't co-occur with simple past (of action/event
verbs):
He still hasn't come.
He has already come.
He hasn't come yet.
370 Particles and illocutionary meanings
*He still came on Friday. (possible with a non-temporal meaning
of still)
*He came already on Friday.
*He didn't come yet on Friday.
On the other hand, events can be dated with the particle only, which, I
suggest, has a separate temporal meaning in addition to those considered
earlier. This temporal meaning can be seen in sentences such as these:
He came back only after he'd squandered all his money.
He came back only when he realised that he wouldn't win.
This sense may be called onlYt (for 'temporal'). It can be spelt out
as follows:
onlYt
it happened at the time when something else happened
it didn't happen before this time
one could think: it would happen before this time
The first component of this explication shows that the event is 'dated',
but it also shows that it is 'dated' by reference to some other event. If we
said, instead, 'it happened at this time', we would fail to account for the
fact that sentences such as the following one are not acceptable (in the
relevant sense):
?He came only at five 0' clock.
(cf. He didn't come until five 0' clock.)
In a sense, therefore, all the temporal particles discussed here are rela-
tive, though in different senses: still, already and yet are relative in the
sense that they don't specify the time of the event, but only situate it
with respect to the time of reference; onlYt is relative in the sense that
while it does specify the time of the event it specifies it not in absolute
terms but in relation to some other event.
To see the meanings of the English temporal particles more clearly I
shall compare them now with their closest counterparts in another
language: Polish, with additional mention of their counterparts in
Yiddish and German.
Polish temporal particles 371
3. Polish temporal particles
3.1. Jut and jeszcze
Leo Rosten (1968:xv) has noted that the following phrases can be
frequently heard in American English, presumably due to Yiddish
influence:
All right already.
This I need yet?
Phrases of the same kind can be heard in the speech of Polish immigrants
in English-speaking countries, presumably for the same reason: Polish,
like Yiddish, has temporal particles which appear to be very close to
already, still, and yet, and which are, consciously or subconsciously,
equated with them in the speakers' minds, whereas in fact the semantic
structures encoded in these Polish or Yiddish particles differ, in some
ways, from those encoded in the English ones (cf. Pasicki 1976). The
Polish particles in question are jUi, often translatable as already,
and jeszcze, often translatable as yet or still. For example:
On jut tam jest.
'He is already (jut) there.'
On jeszcze tam jest.
'He is still (jeszcze) there.'
On jeszcze nie przyszedl.
'He hasn't come yet (jeszcze).'
On jut przyszedl.
'He has already (jut) come.'
But there are also many types of contexts where jut cannot be trans-
lated as already, and where jeszcze cannot be translated as either yet or
still. For example:
Jut czy jeszcze?
'Jui or jeszcze?'
(cf. *Already or still? *Already or yet?)
372 Particles and illocutionary meanings
What the Polish question really means can be stated as follows:
'No more or some more?'
i.e. '(is this) enough or (do you want) some more?'
This is not to imply that jut can always be translated as no more or
jeszeze as some more: jui and jeszeze are, essentially, temporal particles,
and they can be used only in contexts which have something to do with
time (or which can be interpreted as having something to do with time).
Nonetheless, the fact that a virtually context-free sentence such as Jui
ezy jeszcze? means something like 'Some more or no more?' is highly
instructive: it suggests that the two particles include components
crucially involving the notion of 'more'. Similarly, jeszcze would be
translated into English as more in the following context:
Jeszeze raz!
'Once more.' (lit. 'jeszcze once')
Compare, also, the following context, where jeszeze would normally be
translated by even rather than by still:
Zosia jest bardzo ladna, ale Rasia jest jeszeze ladniejsza.
'Sophie is very pretty, but Barbara is even prettier.'
Jeszeze seems to imply here, roughly: 'more; one wouldn't expect to
hear this after what I said before'.
This is, then, the essence of my hypothesis: jut and jeszeze differ
from already and still in so far as the former set is based on the idea
of 'more', whereas the latter is not. As a first approximation, I would
propose an explication along the following lines:
On tam jeszeze jest.
('He is jeszcze there. ')
one could think:
it would not be the same as before this time
it is the same as before this time
it is more of the same
A very similar, virtually identical explication applies also to those
sentences where jeszcze has to be translated by means of even rather
than still:
Zosia jest bardzo ladna, a Rasia jest jeszeze ladniejsza.
('Sophie is very pretty, and Barbara is even prettier. ')
Polish temporal particles 373
(I say: ) Sophie is very pretty
one could think:
1 wouldn't say the same about Barbara as 1 said before (about
Sophie)
1 want to say the same about Barbara
more of the same
On juz tam jest.
('He jui is there. ')
one could think:
it would be the same as before this time
it is not the same as before this time
it is not the same any more
Consider, now, a sentence like the one cited by Rosten:
Jui dobrze.
'Jui all right.' / 'All right already.'
This sentence could be used, for example, in soothing a child:
Here, here. That's enough. It is all right now. ("All right
already." )
The Polish sentence implies:
it is not the same any more
(so there is no need for any further crying)
But while a paraphrase in terms of 'not any more' makes sense here,
one phrased in terms of 'it happened before this time' would not, and
this is, 1 suggest, why already in its standard-English meaning cannot
be used here.
Consider now the other sentence cited by Rosten:
This I need yet?
which can be seen as an attempted English rendering of sentences such
as the following Polish one:
Tego mi jeszcze potrzeba?
'This 1 need jeszcze?'
The sentence is sarcastic and it implies: 'do 1 need this, on top of
everything else?' or 'do 1 need any more (implied: misfortunes, troub-
les)? do 1 need this (trouble)?' Again, the sentence This I need yet?
374 Particles and illocutionary meanings
sounds odd in standard English partly because yet doesn't imply 'more',
which is clearly what the speaker assumes it does. In addition, yet
carries the implication: 'one can think: this will happen', which doesn't
make sense in the present context. But the Polish sentence is felicitous,
because the implications of jeszcze do make sense in this context:
Tego mi jeszcze potrzeba?
(implied: Do I need more (troubles)?)
one could think:
it would not be the same as before this time
it is the same as before this time
it is more of the same (trouble)
I must admit, however, that while the picture presented here appears
to be coherent and to have considerable explanatory power, there are
some other uses of jui and even more of jeszcze, which remain puzzl-
ing and which don't seem to fit this picture. What I have in mind is,
above all, the use of these particles in 'dated' sentences such as the
following ones:
Stalo ~ to jui w pifltek.
'It happened "already" on Friday.'
(i.e. it happened as early as Friday).
Stalo ~ to jeszcze w pifltek.
'It happened "still" on Friday.'
(i.e. it happened as early as Friday).
One puzzling feature of the sentences above is that both jui and jeszcze
seem to mean here the same ('as early as'); and yet the feel of the two
sentences is quite different.
I would suggest that the difference consists here largely in the
speaker's perspective on the event: jui suggests a point of view predating
the event, whereas jeszcze - a present point of view, that is (in this
case) one post-dating it. Roughly speaking, then, jui should be glossed
here not as 'as early as' but as 'as quickly as', whereas jeszcze should
indeed be glossed as 'as early as'. This distinction appears to be sup-
ported by the use of these particles in 'dated' future sentences:
Stanie ~ to jui jutro (?jui w tym tygodniu).
'It will happen jui tomorrow (?jui this week).'
(i.e. this event will come very quickly)
Polish temporal particles 375
Stanie ~ to jeszcze w tym tygodniu (?jeszcze jutro).
'It will happen jeszcze this week' (?jeszcze tomorrow).'
(i.e. it will happen very early)
But why is it that in future dated sentences jut requires precise dating
whereas jeszcze appears to require a reference to periods (which include
'now')? At this stage, I am not able to offer paraphrases which would
account correctly for all aspects of the use of these particles.
It is interesting to note that all the uses of the Polish particles jut and
jeszcze discussed here are shared also by the German particles schon
and noch. Even the two Yiddish-English sentences quoted earlier have
their exact counterparts in German:
All right already!
Schon gut!
This I need yet!
Das fehlt mir noch!
And yet the meanings encapsulated in the German particles schon and
noch must be slightly different from those encoded in the Polish
particles jui and jeszcze, because in some kinds of contexts the former
can be used and the latter not. For example:
a. Du wirst noch sehen!
Jeszcze zobaczysz!
'You will see noch/jeszcze (yet)!'
b. Du wirst schon sehen!
*Jui zobaczysz!
'You will see schon/*jui!'
c. Das wird dir schon leid tun!
Jeszcze tego potalujesz!
'You will regret this schon/jeszcze (yet)!'
d. Das wird dir noch leid tun!
*Jui tego poialujesz!
'You will regret this noch/*jut!'
Finally, as pointed out to me by Gerda Smith (p.c.), in German, noch
and schon can co-occur, whereas jui and jeszcze cannot co-occur, any
more than already and still or yet can:
376 Particles and illocutionary meanings
Das wird dir schon noch leid tun!
*Jeszcze jui tego pozafujesz!
'You will regret this schon noch!'
*You will regret this yet already!
I do not attempt here to construct explications which would account
for both the similarities and the differences between these German
particles and the others discussed here, although the task is most
challenging and tempting.
3.2. Dopiero
Dopiero is another intriguing Polish particle, which seems to lend itself
to diametrically opposed interpretations:
Przyszla dopiero 0 czwartej.
'She came only at four o'clock.' (i.e. late)
Jest dopiero czwarta.
'It is only four o'clock.' (i.e. early)
Similarly:
Dopiero po uplywie kilkunastu minut ukazal ~ we drzwiach
mfody chlopiec. CZeromski) (SIP)
'Only after about fifteen minutes a young boy appeared at the
door.'
Dopiero si6dma. Do dziewifltej jeszcze dwie godziny. (Morcinek)
(SIP)
'It's only seven o'clock. Two more hours are left before nine
o'clock. '
It is very implausible, however, that the same particle would have two
diametrically opposed meanings, 'late' and 'early'. One interpretation
('early') arises in sentences 'dating' the reference time itself; the other
('late') arises in sentences dating the events spoken of.
A unified semantic analysis can be sought along the following lines:
it happened at this time
it didn't happen before this time
one could think: it would happen before this time
Polish temporal particles 377
It is easy to see how the formula sketched above (which is in fact
rather similar to that proposed by Grochowski 1986: 102) can apply to
sentences referring to specific events, for example:
Stalo ~ to dopiero w pifltek.
('It happened dopiero (not before, as late as) on Friday.')
it happened at this time
it didn't happen before this time
one could think: (it would happen) before this time
It is not so easy to see how the same formula could apply to sentences
expressing nothing but time. It seems to me, however, that the same
formula can be seen as applying here, too, if we assume that the speaker
views time as moving in a certain direction. For example:
It is dopiero (no later than) noon.
(i.e. it is noon now)
noon came now
it didn't come before now
one could think: it would come before now
Of course, objectively speaking, noon cannot 'come' sooner or later,
since time always 'moves' at the same speed; but in the speaker's
subjective impression, time can move fast or slowly, and a particular
point in time (e.g. noon), can come sooner or later than one would
(subjectively) expect.
I would suggest, nonetheless, that although all the uses of dopiero
have a common core, polysemy may have to be postulated to account
for its use in those numerical contexts where dopiero clearly implies
'no more'. For example:
Mam dopiero dwoje dzieci.
('I have dopiero two children. ')
I have two children
no more (so far)
one could think: it would be more
one could think: it would happen before this time
it didn't happen before this time
Consider also the common idiomatic expression dopiero co 'just a
moment ago' (lit. 'dopiero what'), where dopiero seems to imply a
very short time. But the explication proposed earlier fits this use of
dopiero, too.
378 Particles and illocutionary meanings
- Kiedy to ~ stalo? ('When did it happen?')
- Dopiero co.
it happened now
(a moment ago, no more)
one could think: it is more
one could think: it happened before this time
it didn't happen before this time
It is easy to see that the Polish particle dopiero corresponds, to some
extent, to that use of the English particle only which I have singled out
as onlYt" In the case of only, however, we could perhaps accommodate
frame-setting sentences such as
It is only four 0' clock.
under the basic, general sense of only ('no more than'). In the case of
dopiero, this cannot be done because dopiero always implies a sequence
of time, even in the purely numerical contexts such as 'dopiero
two children'.
Furthermore, we noted that onlYt cannot be used in sentences dating
events without reference to other events:
?He came only at four 0' clock. (i.e. very late)
To account for this restriction on the use of onlYt' I have postulated for
onlY
t
the following component:
it happened when something else happened
Since no similar restriction applies to dopiero, the relevant component
of dopiero has been phrased differently: 'it happened at this time'.
Polish quantitative particles 379
4. Polish quantitative particles
4.1. Non-approximative particles
4.1.1. Tylko
Tylko corresponds fairly closely to the English particle only and its
meanings can perhaps be represented by the same semantic formulae:
Przyszlo tylko pifC os6b.
'Only five people came.'
tylko
l
=onlYl
no more than this
one could think: it would be more
On tylko robi to na zlosc.
'He's only doing it to annoy us.'
To tylko ja.
'It's only me.'
tylko
2
= onlY2
this
no other than this
one could think:
it would be more than this
Admittedly, we have seen earlier that only and tylko can't always be
substituted for one another. Thus, only can be used in sentences referr-
ing to the passage of time, but tylko can't:
It's only five 0' clock!
*Jest tylko piflta!
If I have, nonetheless, posited the same semantic formulae for tylko and
only, it is on the assumption that only has a third (temporal) meaning,
which tylko doesn't have.
In support of the suggestion that only has one extra meaning, ambigu-
ous sentences such as the following one can be adduced:
He only did it when it was clear that he wouldn't win.
380 Particles and illocutionary meanings
On one interpretation, only implies 'at no other times' and can be trans-
lated into Polish as tylko. On another interpretation, however, (perhaps
a more likely one) it implies 'no earlier than', and cannot be rendered
in Polish as tylko. Instead, a different Polish particle has to be used:
(dopiero):
Zrobil to dopiero wtedy, kiedy bylo jasne, ie nie wygra.
'He did it only (no sooner than) when it was clear that he
wouldn't win.'
4.1.2. At
The particle ai is, in some ways, an opposite of tylko. If tylko implies
that 'one could think that it would be more', ai implies that 'one could
think that it would be less'. Often, to replace at with tylko one would
have to replace, at the same time, much with little, or for a long time with
for a short time, and vice versa.
Wizyta trwala dlugo, bo ai do p6znego wieczora. (Orzeszkowa)
(SJP)
'The visit lasted for a long time, in fact right up to the late
evening.'
We could try, therefore, to posit for ai a meaning which could be a
mirror image of that posited for tylko:
no less than this
one could think: it would be less
However, although symmetrical patterns of this kind are always appeal-
ing to the analyst, symmetry should not be cherished over and above
empirical accuracy. It should be pointed out, therefore, that there are
contexts where the differences in the behaviour of ai and tylko can't be
explained in terms of the neat formulae sketched in above. In particular,
these formulae don't explain why ai can't very well co-occur with an
imperative, whereas tylko can:
Daj mi tylko p i ~
'Give me only five!'
?Daj mi at p i ~
'Give me as many as five!'
Polish quantitative particles 381
I think that this difference between at and tylko may be connected
with the fact that at seems more emphathic, more expressive, more, so
to speak, 'surprised' than tylko. Tylko, like only, is fully compatible with
a purely descriptive, objective tone. But at isn't. To account for this
fact, it seems more appropriate to formulate the relevant component of
at as follows:
one would have thought: it would be less
The symmetrical component: 'one would have thought: it would be
more' would not be suitable for tylko, or for only, which don't sound
'surprised' but merely 'cautious'. I posit, therefore, the following
semantic formula:
at
this
not less than this
this is much
one would have thought: it would be less
English has no particle corresponding to at. The closest to it is the
particle-like expression as many as (as much as), but at is not similarly
restricted to numbers and measurable quantities. For example, one
cannot say in English:
?The visit lasted as long as up to the late evening.
whereas as we have seen, at can very well be used in this kind of
context.
4.1.3.
combines the roles of a particle and of a conjunction. It
occurs either in concessive clauses or in temporal ones, with slightly
different force in each case. The concessive use can be illustrated by
the following sentence:
Nie skartyl na choc od paru dni sypial zaled-
vvie po godzin na (L. Bartelski) (SIP)
'He didn't complain of fatigue, although for the last few days he
had been sleeping only (zaledvvie) a few hours per day.'
382 Particles and illocutionary meanings
The semantic contribution of zaledwie in such sentences appears to
be this:
zaledwie
this
no more (time) than this
this is not much
one would think: it would be more
Usually, the 'substance' whose quantity is being estimated is time. I have
put the word time in parentheses, however, to indicate that sentences
with zaledwie referring to other 'substances' can also be encountered
and can be accepted (at least by some native speakers):
Dala mu zaledwie p i ~ dolarow (ale on s i ~ nie obrazil).
She gave him only five dollars (but he didn't take offense).
The concessive context improves, in my judgement, the acceptability
of the sentence, but again there seems to be some variation in this area.
In temporal clauses, non-concessive zaledwie signals an even smaller
amount of time ('next to nothing', 'hardly any') than it does in conces-
sive ones.
Zaledwie wzifllem pioro do r ~ k i zapukal ktos do drzwi. (Jan
Lam) (SJP)
'I had hardly picked up my pen when somebody knocked on the
door'.
In sentences of this kind zaledwie signals an almost direct succession
of events. It also conveys the idea that the quick succession of events
is unexpected. An explication of the temporal meaning of zaledwie
(zaledwie
2
) is not attempted here.
4.1.4. Ledwie
Ledwie is closely related semantically, as well as morphologically, to
zaledwie, and both can often be rendered in English as hardly. Yet
the two are not always interchangeable, and the semantic contribution
of each is distinct. In particular, ledwie is not restricted to time, in
the way zaledwie is (in the modem usage). The use of ledwie can be
illustrated with the following sentences:
Polish quantitative particles 383
sie
'I just managed to squeeze in.'
Ledwie go poznalam.
'I could barely recognise him.'
Ledwie mi wystarczylo.
'I had barely enough.'
Ledwie implies that there was 'just enough' of something for X to
happen, and that had there been even a tiny bit less of it, X wouldn't
have happened. For example, had the friend changed just a little more
than he had the speaker wouldn't have recognised him. Had the
speaker had even a tiny bit less money than he had it wouldn't have
been enough. Had there been even a very little less room than there
was the speaker wouldn't have managed to squeeze in. And so on.
The link with zaledwie, especially with zaledwie
2
, is evident. In the
case of zaledwie
2
the margin of time is 'next to nothing', and in the case
of ledwie the margin of whatever is 'next to nothing'. It can even be
the margin of time, as in the sentence:
Ledwie zdflzylam.
'I just made it.' (i.e. 'I barely made it. ')
But ledwie can be used in a simple sentence, where it doesn't refer to
the relationship between two events X and Y. Zaledwie
2
can't be used
like that:
*Zaledwie zdflzylam.
Zaledwie
2
stresses the minimal difference in time between two events.
By contrast, ledwie stresses the minimal 'distance' between the happen-
ing and the non-happening of an event. To say that something ledwie
happened is close to saying that it nearly (almost) didn't happen.
Ledwie zdflzylam na poci{lg.
'I just made it to the train.' = 'I nearly missed the train.'
Furthermore, ledwie seems to imply that if the event in question didn't
happen it would have been unfortunate, undesirable. For example, one
can't say:
*Ledwie spoiniles.
'You just missed it.'
384 Particles and illocutionary meanings
as one can say:
Ledwie zdflzyles.
'You just made it.'
In fact, ledwie typically co-occurs with 'verbs of success', such as
zdpzyc 'make it on time', wystarczyc 'suffice', zdac 'pass an exam,
successfully', udac ~ 'manage', wytrzymac 'bear' or wycipgnpc ~ z
czegos 'pull through'.
The following semantic formula can be proposed for ledwie:
ledwie
if something was no more than a little different
X would have happened
this would be bad
4.2. Polish approximative particles
4.2.1. 0 malo nie
The expression 0 malo nie (literally 'by little not', i.e. 'but for a little
margin it wouldn't have happened') is frequently interchangeable with
ledwie - not in the same sentences but in the same situations:
Ledwie zdflzylam na poci{lg.
'I barely made it to the train.'
o malo ~ nie sp6inilam na pociflg.
'I nearly missed the train.'
However, ledwie can be applied to an ongoing situation as well as to
an event, whereas 0 malo nie (like nearly) applies only to momentary
events.
Ledwie mi wystarcza.
'I can barely make ends meet.'
*0 maio mi nie wystarcza.
'I nearly can't make ends meet.'
To account for this 'momentary' implication of the expression 0 malo
nie 1 have formulated its second component as 'X would have happened
at that moment'.
Polish quantitative particles 385
Finally, 0 malo nie shares the 'evaluative' character of ledwie (both
imply that a disaster, major or minor, has barely been averted). Hence
the need for the component 'this would be bad'. The overall formula
for 0 malo nie might read:
o malo nie
if it was no more than a little more
X would have happened at that moment
this would be bad
4.2.2. Niemal and prawie
Niemal (nie+mal, lit. 'not+little') may seem to be just a variant of 0
malo nie, alongside perhaps, omal and nieomal ('not+by+little'). In
fact, however, niemal differs from both 0 malo nie and omal (if not
from nieomal) in several important ways. First, niemal refers to static
states of affairs rather than to events or other dynamic situations.
Thus, one can say:
Odziez pOS10W byla tak skromna, ie graniczyta niemal z
ubostwem. (Mieczyslaw Gomulicki) (SIP)
'The clothing of the envoys was so modest that it almost (niemal)
bordered on poverty.'
But one could not say:
*Niemal ~ spoznilam na poci{lg.
'I nearly (niemal) missed the train.'
Furthermore, niemal is free of the negative connotations of omal and 0
malo nie, and can be used in positive contexts:
Byla bardzo urocza w swym niemal dziecinstwie, a juz
kobiecosci. (Zeromski) (SIP)
'She seemed almost (niemal) a little girl, and already a woman,
and was very appealing.'
It should be noted that despite what its morphology suggests, niemal
can refer to a minimal difference in quality, as well as to a minimal
difference in quantity. In other words, what is at issue is not so much the
idea that 'if it was a little more then it would be true to say X' as the idea
386 Particles and illocutionary meanings
that 'if it was no more than a little different then it would be true to
say X'. For example, one can say:
S(l niemal jednakowe.
'they are nearly (niemal) identical.'
Purely quantitative particles, such as blisko, cannot be used like that
(see the section below).
Prawie is extremely close to niemal, and the two are almost inter-
changeable. The difference between them may seem purely stylistic,
niemal being slightly less colloquial. Nonetheless even here the
synonymy is probably not complete. There are contexts where prawie
sounds much better than niemal, and evidently not for stylistic reasons.
For example:
Jui prawie (?niemal) skonczylam.
'I have already almost finished.'
Dala mu prawie (?niemal) sto dolarow.
'She gave him almost one hundred dollars.'
Przyszlo prawie (??niemal) 20 os6b.
'Almost 20 people came.'
The fact that niemal is least appropriate in numerical and otherwise
factual contexts suggests that it has something to do with the speaker's
evaluation. Prawie is not similarly restricted. To account for this (slight)
difference 1 would phrase the explication of prawie in terms of what
'one can say', in contrast to the explication of niemal, which has been
phrased in terms of what 'I want to say'. As a first approximation, then,
the following can be proposed:
prawie
one can't say this
if something was no more than a little different
one could say this
niemal
I don't want to say this
if something was no more than a little different
1 would say this
However, if these explications were entirely correct, prawie would be
identical in meaning to the English particle almost. But in fact, prawie
Polish quantitative particles 387
and almost cannot be identical in meaning, since prawie, unlike almost,
cannot be used in sentences which would be made patently false by
the deletion of the particle; for example:
She almost killed him.
To translate this sentence into Polish, one would have to use 0 malo nie,
not prawie. Prawie can only be used in sentences where the deletion
of the particle would change the sentence from true to 'almost true', not
to patently false; and the same applies to niemal.
To account for this important difference between prawie and almost
I would propose that their explications can be differentiated as follows:
almost
one can't say this
if something was no more than a little different
one could say this
prawie
one can't say this
if one said this
it would be no more than a little different from what one
can say
For example, the sentence:
Sp prawie jednakowe.
'They are almost identical.'
implies that while one cannot say 'they are identical', if one did say
this it would be no more than a little different from what is true. One
cannot say, however:
*Prawie go zabila.
'She prawie killed him.'
because this would imply that in saying zabiia go ('she killed him') one
would be saying something no more than a little different from what
is true (that is, from what one can say). The English sentence:
She almost killed him.
is felicitous, because it doesn't imply this. Rather, it implies that if
something happened that was no more than a little different from
what did happen, it would be true to say 'she killed him'.
388 Particles and illocutionary meanings
4.2.3. Blisko
Blisko (literally 'nearly') is indeed very similar to the English particle
nearly, and differs from prawie in much the same way as nearly
differs from almost. However, blisko differs from nearly, as prawie
differs from almost, in not being able to be used in sentences when the
deletion of the particle would make the sentence glaringly false:
I nearly died of exhaustion.
*Blisko umarlem z wyczerpania.
As for the relationship between blisko and prawie, blisko has a
narrower range of use. In quantitative contexts, both blisko and prawie
can be used:
Przyszlo blisko (prawie) sto os6b.
'Nearly/almost one hundred people came.'
In qualitative contexts, however, prawie is still appropriate, but blisko
is not:
Sfl prawie (*blisko) takie same.
'They are almost (nearly) identical.'
As the last example shows, blisko differs in this respect from
nearly, as well as from prawie and almost. Consider also the following
contrasts:
Prawie (*blisko) nikt nie przyszedl.
'Almost nobody came.'
Jestesmy prawie (*blisko) na miejscu.
'We are almost (/nearly) there.'
'Jut prawie (*blisko) skonczylismy.
'We have almost (/nearly) finished.
Evidently, blisko should be assigned a semantic formula phrased in
terms of 'a little less' rather than just 'a little different'. I would propose
the following:
blisko
it is not this
it is a little less than this
Conclusion 389
if one said: 'it is this'
it would be no more than a little different from what one
can say
5. Conclusion
I hope to have shown that even subtle nuances of the meaning of
particles can be captured in paraphrases whose empirical adequacy can
be verified by substitution in context, as envisaged by Leibniz.
The data discussed in the present chapter suggest, it seems to me,
that particles offer a particularly fruitful field for structural semantic
analysis. The contrast between the volume of literature which has been
devoted to just a few, always the same, semantic fields, such as colours
and kinship, and the paucity of literature devoted to other highly struc-
tured fields, such as particles, seems to me quite remarkable.
Examining the citations for a number of particles in a large historical
dictionary, such as SIP, I am also struck by the clarity with which
such a body of citations reflects on-going semantic change, and by the
relative ease with which semantic mutations in the area of particles
can be captured in the form of paraphrases of the kind employed here.
I conclude that far from being one of the most idiosyncratic and
'fuzzy' areas of the lexicon, particles offer in fact an excellent example
of a highly structured semantic domain, and constitute a rewarding field
for methodological experimentation.
Chapter 10
Boys will be boys:
even 'truisllls' are culture-specific
1. The meaning of tautologies
What is the meaning of English sentences like Boys will be boys, Boys
are boys, Kids are kids, or Business is business? One popular answer
is that such sentences (roughly, X is X) are patent tautologies, and so
necessarily true. Their meaning - which is identified with their 'logical
form' - can be informally stated as follows: 'For every entity of which
it is true to say that it is a boy, it is true to say that it is a boy.' ("A
tautology is a symbolic sentence whose truth value is T with respect to
every possible assignment. For instance, P -> P is a tautology." Kalish
- Montague 1964:74.) That is, it is assumed that the syntactic structure
N is N is exactly equivalent to the logical formula p = p. In fact, of
course, sentences of this kind convey more. As Levinson (1983:125) puts
it, the implication is: "That's the kind of unruly behaviour you would
expect from boys." But, according to the pragmatic explanation, this
implication is a conversational implicature, calculable from Grice's
(1975:45) maxim of Quantity: roughly, "Make your contribution no
more, and no less, informative than required."
I want to argue against this account, and against the whole vision
of linguistics which goes with it. The question of how to interpret
sentences such as Boys will be boys may seem minor and unimportant.
I think, however, that the consequences of one's stand on this point
are far-reaching; they determine one's entire idea of linguistics, its
boundaries, its capacities, and its responsibilities.
In choosing the expression Boys will be boys as the title of this
chapter, I do not wish to imply that this is a paradigm example. It is a
frozen expression, whereas the chapter is concerned above all with the
productive pattern NP
i
be NPj" Still, the 'proverb' has been frequently
discussed in connection with linguistic 'tautologies' and Gricean max-
ims, and it has a greater evocative and mnemonic value than, say, A
promise is a promise or A man is a man. Although there are of course
392 Boys will be boys: even 'truisms' are culture-specific
other types of tautological utterances, the linguistic literature on 'tautolo-
gies' refers largely to sentences of the form X is X.
While the construction type considered in this chapter is a minor one,
it should perhaps be noted that tautological constructions are not at all
rare in colloquial English speech. For example, in three successive
episodes (of 25 minutes each) of the American television program 'Be-
witched', I counted no fewer than seven occurrences. Some cultural
aspects of what things are so 'obviously' true that they can be stated
in the form of a 'tautology' will also be considered here.
1.1. Gricean maxims: universal or language-specific?
My first objection to the Gricean account of 'tautologies' is that it
suggests that the import and use of such constructions should be
calculable from some universal, language-independent principles. In
fact, however, it is not. For one thing, some English 'tautological
constructions' have no literal counterparts which can be used or inter-
preted in many other languages. Some English tautological constructions
do have literal counterparts in other languages, which are used, how-
ever, with a different communicative import. This suggests that, in each
case, the communicative import is conventionally encoded in a given
construction, and is not calculable from any language-independent
pragmatic maxims.
In saying this, I am not arguing against the validity or the significance
of language-independent pragmatic maxims like those posited by Grice.
I am arguing only against the use to which such maxims have been put
in much current literature on linguistic pragmatics, and in particular
against attempts to explain the use of English 'tautological construc-
tions' exclusively in terms of universal pragmatic principles, as 'radical
pragmaticists' try to do (cf. for example Cole 1981). Instead, I am
advocating what might be called a 'radically semantic' approach to the
task: I argue that the constructions in question have a language-specific
meaning, and that this meaning must be spelt out in appropriate semantic
representations.
For example, sentences of the kind adduced in the title of this
chapter. are not used in French, German, or Russian. French sentences
like Lesgarr;ons sont les (des?) r ~ o n s ' (The) boys are (the) boys', or
Les garr;ons seront les (des) r ~ o n s 'The boys will be (the) boys',
would be simply incomprehensible to French speakers. Conceivably,
The meaning of tautologies 393
one could be understood if one said Les gar(;ons seront toujours les
(des) garc;ons 'Boys will always be boys' (cf. Bally 1952: 17), but even
this would be puzzling. Similarly, in German one would not say Knaben
sind Knaben 'Boys are boys', or Knaben werden Knaben sein 'Boys will
be boys.' If foreigners did use such sentences, wishing to convey the
messages of their literal English counterparts, they might not be under-
stood, as the 'implicature' of the tautological constructions in question
would not be understood. The sentence Knaben bleiben (immer) Knaben
'Boys remain (always) boys' is more readily interpretable; but here it is
doubtful if its literal English equivalent would be similarly interpreted.
Russian has three copula constructions: one with eto, one with est',
and one with zero, but none of these would be used to translate the
relevant English sentences:
??Mal' tiki eto/est' mal' tiki.
'Boys are boys.'
??Mal'ciki mal'ciki.
'Boys (are) boys.'
??Mal'ciki budut mal'tiki (mal'tikami).
'Boys will be boys.'
??Deti eto/est' deti.
'Kids are kids.'
To translate the English sentences, one would use a particle:
(Cego ty xoces'?) Oni ze mal' ciki.
(what you want) they PRT boys
'(What do you expect?) They are boys.'
(Cego ty xotel?) Oni ze deti.
(what you wanted) they PRT children
'(What do you want?) They are children.'
It should be pointed out, however, that the Russian particle ie is not
used specifically for the purpose expressed by English sentences like
Kids are kids. Rather, it appears in a variety of constructions, to
indicate, roughly, that something should be obvious to the addressee. It
is thus similar to the German particle doch (cf. Rath 1975; Sekiguchi
1977), though the two are by no means fully equivalent.
However, 'tautological constructions' with action nouns do exist in
Russian. For example, one can say, as Bulat Okudzava does in a popular
394 Boys will be boys: even 'truisms' are culture-specific
song, Rabota est' rabota, rabota est' vsegda, 'Work is work, there is
always work.' One can also say Vojna est' vojna, 'War is war', but the
intended meaning of the Russian expression, and its range of use, would
be different from those of its literal English counterparts. For example,
in Grossman's novel Zizn' i sud' ba 'Life and fate', one of the heroes
uses this expression to express his determination to fight fiercely:
This isn't the moment for me to start talking about internationalism or
class conscience. What matters is to mobilise the fury of the masses against
the enemy.... There's no place for Christian humanitarianism now....
now the Germans have attacked the homeland of workers and peasants.
War's war! (Vojna est' vojna!) They deserve what they get. (Grossman
1980:153; 1985:237)
In the English translation above, the tautological expression War's war
sounds somewhat incongruous, because this is not how this expression
is used in English (see section 2.1 below).
Interestingly, in English one can say That's life, but not *Life is
life. On the other hand, in Russian, where the tautological pattern in
question affirms that one has to do something (rather than put up with
something), the expression Zizn' est' zizn', 'Life is life', is perfectly
possible. For example:
The old woman sitting next to him, the mother of the wife he had loved and
now lost forever, kissed him on the head and said: 'It doesn't matter,
Stepan my dear, it doesn't matter. It's life. CZizn' est' zizn' .)' (Grossman
1980:605; 1985:867)
The Russian original implies that 'life has to go on', but the English
translation implies something different: a sober, 'realistic', worldly-wise
attitude to life.
Furthermore, in French one can hardly say La guerre est la guerre
'War is war'; to express a similar idea, one would say C' est la guerre
'That's war', just as one says C' est la vie 'That's life'; or La guerre,
c' est la guerre, literally 'The war, that's the war'.
To widen the sphere of comparison, in Polish one cannot say *Chlopcy
to ~ d l chlopcy 'Boys will be boys', or ??Chlopcy to (S{l) chlopcy
'Boys are boys'; however, one can say (Jednak) co Paryi to Paryz
'(However) what (is) Paris this (is) Paris', or (Jednak) co Europa to
Europa '(However) what (is) Europe this (is) Europe.' I invite the
'radical pragmaticists' to work out the communicative import of this
construction. For those readers who acknowledge that they cannot
The meaning of tautologies 395
work it out for themselves, here is an approximate answer. The Co X to
X construction ('What is X is X') implies that there is something
uniquely good about X, and that the speaker feels he must admit it. (No
matter what virtue one may find in other comparable things, one must
admit that X is uniquely good.) The construction is used with respect
to referents which are well known, and which are widely regarded as
uniquely good. Given traditional Polish attitudes, Paris qualifies par
excellence; but any other well known symbol could also be so men-
tioned, in grudging recognition of its widely-acclaimed superiority.
Hence, for example, the following contrasts:
(Jednak) co a l ~ s a to a l ~ s a
??(Jednak) co Barbara to Barbara.
(Jednak) co rodzina to rodzina. [rodzina 'family']
??(Jednak) co kuzyni to kuzyni. [kuzyni 'distant cousins']
The Polish construction Co X to X can be seen as related, in certain
respects, to the English pattern illustrated by the sentence used recently
as an opening statement in a television interview by a Ku Klux Klan
leader: White is white. In such sentences, the speaker stresses the unique
quality of something which must be accepted because it cannot be ex-
pected to change. But qualities like 'whiteness' are seen as belonging to
certain contrastive sets; so, by stressing their uniqueness, the speaker
emphasises the irreducible difference between the members of the set.
The uniqueness is not interpreted here as superiority, but merely as a
reason for an irreducible contrast, which must be accepted as such. The
attitude encoded in this construction is highlighted in the familiar
passage from Kipling's 'Ballad of East and West':
East is East and West is West
and never the twain shall meet.
In the film A passage to India, based on E. M. Forster's novel, Kipling's
saying is used in a reduced version, as East is East; this version
seems parallel to White is white (with the second member of the contras-
tive set being only implied.)
In the Polish construction, no finite sets of comparable elements are
implied, and so the notion of uniqueness is not tantamount to one of
irreducible difference; instead, it is tantamount to a notion of irreducible
superiority (with respect to any conceivable competitors).
396 Boys will be boys: even 'truisms' are culture-specific
It is interesting to note that a different but somewhat related 'tautol-
ogy of admiration' exists in Korean; one can say Apoci-ja apoci ida
'Father is father' (where -ja means, roughly, 'certainly'), or Colsu-ja
Colsu ida 'Colsu [a person's name] is Colsu', to express one's admira-
tion for the person referred to. However, Korean also has a tautological
construction which is used to express disapproval. For example, seeing
a person crying without any good reason, one might say to him, with
a scornful laugh, No-to no ida 'You are you' (where -to means, roughly,
'also'). In English (or in Polish), in these circumstances, one would
have to say something quite different; for example, It's just like you
(that's you allover) - crying for no reason at all. (lowe the informa-
tion on Korean to an unpublished paper by Gi-hyun Shin.)
As a final example, consider the following Japanese sentences, which
come from unpublished papers by Itsuo Harasawa (1985) and Deborah
Field (1988):
Makeru toki wa makeru yo.
'When (I) lose (a game), (I) lose (it).'
Kare datte, kekkonsuru toki wa kekkonsuru yo.
'Even he, when (he) marries, (he) marries.'
Okoru toki wa okoru.
'When (he) gets angry (he) gets angry.'
Can native speakers of English work out the 'implicatures' of such
sentences? Several speakers of English whom I have asked have offered
the following: 'When he loses the game, he loses it badly'; 'When he
gets married, he will get married in a spectacular way.' 'When he gets
angry, he gets really very angry'. Clearly, these guesses are based on the
interpretation of English tautologies like the following:
"Four cakes! Gee!" said fern.
"When we give a party, we give a party," said Susan grandly.
(Montgomery 1980:73)
But in fact these guesses turn out to be incorrect. According to
Harasawa, the Japanese construction really means that something
regarded as quite impossible is actually possible. It would seem that this
meaning is language-specific, and cannot be calculated solely on the
basis of any Gricean maxims (or 'post-Gricean' ones, cf. Atlas 1984).
The meaning of tautologies 397
Again, I do not wish to deny that the various 'tautological construc-
tions' used in different languages have something in common, and that
they may be partly explained in terms of some language-independent
principles such as Grice's maxim of quantity. I would insist, however,
that their use cannot be fully accounted for in such terms. Of course,
most facts of grammar can be partly explained in terms of language-
independent principles - functional, perceptual, logical etc. If
linguistic descriptions of particular languages were to be purged of
everything that could be so dealt with, then little would be left for
the linguist to write about.
When philosophers write about the constructions of their native
language as if their use were fully determined by the general laws of
human reason, or by 'normal' rules of conversational cooperation, they
can perhaps be excused: it is not the job of philosophers to compare
different languages, or to be aware of both the differences and the
similarities between them. But when reputable linguists eagerly and
enthusiastically adopt the philosophers' illusions, the situation begins to
look like a historical aberration. One can understand Givan's exaspera-
tion on this score, expressed in the following outburst (whether or not
one agrees with his proposed solution):
Nobody would deny the stimulating effect that a first reading of Grice's
'Logic and conversation' [1975] may produce. But to base an entire
boomlet, indeed a fad, on this rather limited construction of the pragmatic
agenda in terms of Grice's 'maxims', from which all else is presumably
derived as deus ex machina, is the climax of in-group folly.' (Giv6n 1983:
154) (Cf. also Boguslawski 1981a.)
1.2. Problems in interpreting implicatures
According to Levinson (1983:124), among others, a sentence like Boys
are boys is necessarily true. I dispute the validity of this statement,
which reflects a mistaken belief that the sentence under discussion
is factual. It is clearly not: it expresses a certain attitude, and attitudes
can hardly be called 'true' or 'false'. Roughly speaking, it is a call for
tolerance, an injunction; and it is no more 'true' than the Ten Command-
ments, or maxims like Time is money or The early bird gets the worm.
Since the attitude in question cannot be fully worked out on the basis
of any language-independent principles, it must be regarded as the
language-specific meaning of the sentence in question. It is the responsi-
398 Boys will be boys: even 'truisms' are culture-specific
bility of the linguist to identify this construction, and to spell out the
meaning encoded in it. To the best of my knowledge, neither task has
been undertaken by any of the numerous linguists who have written
about tautological sentences in a Gricean vein.
A foreigner trying to learn English will have to be taught some rules
for correctly using various tautological constructions. It would be
perfidious in the extreme merely to draw his attention to the Gricean
maxims, and then to leave it to him to work out for himself the
permissible range of use of these constructions.
An interpretive formula like that quoted above - by which Boys are
boys implicates something like 'That's the kind of unruly behaviour
you would expect from boys' - says both too much and too little.
On the one hand, the epithet 'unruly' is over-specific and arbitrary. On
the other hand, the formula is not specific enough, in not spelling out
the speaker's tolerant and indulgent attitude. Consequently, it doesn't
predict that the following sentences are odd (the behaviour in question
being too 'bad' to be treated indulgently):
?Sadists are sadists.
?Rapists are rapists.
?Nazis are Nazis.
The subtle difference in attitude between the sentence Boys will be boys
and sentences such as Business is business is implicitly acknowledged in
the glosses offered by Cowie's (1976) Dictionary of current idiomatic
English. According to this dictionary, the former saying is offered as an
'excuse', whereas the latter constitutes an attempt at 'justifying oneself'
(or, indeed, others). I think this differentiation reflects a subtle insight,
of the kind that radical pragmaticists or other Griceans prevent them-
selves from being able to capture.
The tolerant and indulgent attitude is even more transparent in the
pattern with the modal will:
Boys will be boys.
Students will be students.
*Wars will be wars.
*Business will be business.
The pattern with will implies that the nature of the human beings in
question cannot be repressed; one shouldn't try to change boys (etc.)
because it won't work anyway: whatever one may do, boys will still
behave like boys, so the wisest course is just to let them be.
The meaning of tautologies 399
Levinson's formula is not nearly as unsatisfactory as many other
similar formulae offered as statements of various 'conversational
implicatures', but it still illustrates the detrimental effect of the 'radical
pragmatics' doctrine upon the analysis: when linguists assume that the
communicative import of a construction is a matter of implicature,
rather than meaning, they then seem not to care about how, exactly, this
import is to be stated. In a sense, radical pragmatists can't afford the
luxury of stating the differences in the communicative import between
different tautological constructions, either across language boundaries or
within a single language. By assuming an absence of meaning (other
than that equivalent to the 'logical form'), they are also forced to assume
an absence of polysemy; thus they must either posit a single formula
which may be too broad to predict the exact range of use, or else must
be rather vague about the details of the alleged implicature. If a method-
ology requires us to devise a semantic formula (or a few formulae) for
each construction, it forces us to be explicit and precise. As a result,
differences as well as similarities come to light which would otherwise
be missed - or which may be vaguely attributed to differences in
'context', with no precise generalisations being offered or sought. An
illustration of this last failing is provided in the next section.
Before we turn to this illustration, however, let it be pointed out that
an eagerness to make the Gricean maxims work at all costs can be also
detrimental to the analysis from a purely formal point of view. Sentences
such as Boys will be boys provide a good case in point. For isn't it
strange that several writers who have discussed such sentences have
followed Grice (1975:52) in focussing exclusively on the less idiomatic
variant Boys are boys and have failed even to note the existence of the
variant with will? It is hard not to regard this omission as 'purposeful',
on an unconscious level. Boys are boys looks like a tautology, so it
may seem to be explainable in terms of the 'maxim of quantity'. But
Boys will be boys is not a tautology: normally people assume that boys
will not always be boys (but will become men). How is it then, that
such an obvious non-tautology can have almost the same 'conversational
implicature' as a 'tautology' (Boys are boys)? Both the form and the use
of the version with will seem to undermine the validity of the Gricean
account. As a result, this version is simply left out of the account alto-
gether! I am certainly not suggesting that this is a case of intellectual
dishonesty, but I am suggesting that the omission in question is a
Freudian omission, so to speak.
400 Boys will be boys: even 'truisms' are culture-specific
1.3. Context as an excuse for analytical failure
It will be useful to take, as a starting point, a longish passage from
Levinson:
The uttering of simple and obvious tautologies should, in principle, have
absolutely no communicative import. However, utterances of (38)-(40) and
the like can in fact convey a great deal:
(38) War is war.
(39) Either John will come or he won't.
(40) If he does it, he does it.
Note that these, by virtue of their logical forms (respectively:
Vx [W(x) -> W(x)]; P V p; p -> p) are necessarily true; ergo they
share the same truth conditions, and the differences we feel to lie between
them, as well as their communicative import, must be almost entirely due
to their pragmatic implications. An account of how they come to have
communicative significance, and different communicative significances,
can be given in terms of the flouting of the maxim of Quantity. Since this
requires that speakers be informative, the asserting of tautologies blatantly
violates it. Therefore, if the assumption that the speaker is actually co-
operating is to be preserved, some informative inference must be made.
Thus in the case of (38) it might be 'Terrible things always happen in war,
that's its nature and it's no good lamenting that particular disaster'; in
the case of (39) it might be 'Calm down, there's no point in worrying
about whether he's going to come because there's nothing we can do about
it'; and in the case of (40) it might be 'It's no concern of ours.' Clearly
these share a dismissive or topic-closing quality, but the details of what is
implicated will depend upon the particular context of utterance. (Inciden-
tally, exactly how the appropriate implicatures in these cases are to be
predicted remains quite unclear, although the maxim of Relevance would
presumably playa crucial role.) (Levinson 1983: 110-111)
To Levinson's credit, he does point out that "exactly how the
appropriate implicatures in these cases are to be predicted remains
quite unclear". He still pins his hopes, however, on the context of the
utterance. I believe that such hopes are bound to be disappointed. Utter-
ances like War is war or Boys will be boys are remarkably context-
independent in their force, as Levinson himself tacitly recognises by
trying to spell out their 'implicatures' without asking the reader to
imagine any particular context. Grice too (1975:52) discusses the
'conversational implicatures' of such sentences without invoking any
particular context. In fact, various dictionaries of 'sayings', proverbs,
The meaning. of tautologies 401
and idiomatic phrases have often listed such expressions, and have also
explained their meanings, as best they could, without appealing to any
particular contexts (cf. Bartlett 1980; Cowie 1976; Stevenson 1949).
The lexical meaning of the word war may indeed influence the
'implicature' of the saying War is war; but this doesn't mean that the
construction itself is similar in meaning to Boys are (will be) boys.
Both constructions signal a 'philosophical' attitude, a kind of acceptance
of bad things which cannot be prevented from happening; but the
plural-human construction also signals an indulgent attitude, while the
singular-abstract one does not.
The element will, which can show up in the sentence as an alternant
of are, is not accidental either: it reflects both a reference to the future
(you can predict how boys will behave from their nature) and to a
characteristic human 'obstinacy' or a tendency to persist in behaving
in certain ways, no matter how undesirable from other people's point
of view. This will is related to the future will; but it is also related
(synchronically) to the noun will and to the modal used in sentences
referring to habitual (stubborn) 'undesirable' behaviour (He will smoke
while we're still eating).
It is worth recalling in this connection Jespersen's (1965) remarks on
the use of will, as well as his examples illustrating the semantic links
between human will, human nature, habitual behaviour, and a tendency
to stubborn persistence in what might seem foolish:
Another connected transition is a consequence of the fact that what one
does willingly, one is apt to do frequently. Hence will ... comes to be the
expression of a habit, especially a habit which is a consequence of one's
character or natural disposition ... In the present tense, it does not seem
usual in the first person ... , in the second it is often emotionally coloured:
'You will smoke all day long - and then complain of a sore throat!' ... If
will is emphasised, obstinacy may be meant: Gammer 102 fooles will be
fooles styli! boys will be boys... (Jespersen 1965,4:240-241)
Once 'human' tautologies, such as Boys are boys, are distinguished
from abstract ones like War is war, the appearance of will in some
'tautological' sentences, but not others, begins to make sense.
Turning now to sentences like Either John will come or he won't, I
would agree with the gist of Levinson's interpretation ('Calm down,
there is no point in worrying about whether he's going to come because
there is nothing we can do about it. '). But if it is stated in these words,
one can't see how this particular 'implicature' is related to the 'implica-
402 Boys will be boys: even 'truisms' are culture-specific
tures' of the other tautological constructions considered. If one says
that all the different tautological constructions express a certain
philosophical acceptance of undesirable events, one will be more or
less right - but only more or less. For example, the nuance of
tolerance/indulgence conveyed by Boys are boys and Boys will be boys
is lost. (No such nuance is conveyed in 'fatalistic sentences' like Either
John will come or he won't.)
One might try to defend the claim that the differences between the
various 'tautological' constructions can be predicted from the context
in the following way. If the sentence refers to future events, the implica-
tion is that future events are unknown and must remain unknown (for
the present), so there is no point in speculating about them. If the
sentence refers to past events, the implication is that what has already
happened can't be changed, so it is not worth dwelling on. If the refer-
ence is not to any past event, but specifically to a past action by the
speaker himself, the implication may be that of determination and refusal
to change one's position, as in Pilate's utterance: What I have written I
have written. If a sentence has a generic meaning, and predicates some-
thing about a 'species', then the implication will be that the nature of an
individual is determined in some respects by the nature of the species,
and so cannot be changed - and moreover that it should be excused, to
the extent that 'bad' behaviour of an individual is determined by the
nature of the species.
Indeed, the Spanish sentence Que sera, sera conveys not only a
philosophical acceptance of what cannot be changed, but also a dismissal
of useless speculations about the future. The Polish saying Co bylo to
bylo 'What was (has been), this was (has been)' implies that past
events cannot be changed, so there is no point in dwelling on them. In
fact, this saying has a more explicit version, which has the status of
a proverb: Co bylo a nie jest nie pisze sit? w rejestr 'What has been but
is not shouldn't be put on record.'
Nevertheless, and despite appearances, the exact meaning of sentences
which express a fatalistic attitude to future events (like the Spanish Que
sera, sera) can also vary from language to language. For example, the
best rendering of the Spanish sentence into Polish would be Co rna bye to
~ z i e 'What is to be, that will be', and not Co e z i ~ to e z i ~ 'What
will be will be'. The latter sentence is used in Polish to express one's
determination to act, regardless of possible negative consequences;
thus it can be used by a soldier before a battle or by a student before an
exam. Characteristically, it is often followed by the proverb Raz kozie
English nominal tautologies: semantic representations 403
smierc 'A goat has to die once', or by the saying Co ma wisiec nie
utonie 'What is meant to hang will not drown', both of which express
a similar sentiment of reckless determination. The Spanish sentence can
be used by an idle person who doesn't want to worry about the future,
but its literal Polish equivalent would not be so used.
Facts of this kind show, it seems to me, that the use of 'tautological
constructions' and 'tautological sayings' in different languages is partly
conventional and language-specific - even though it can also be ex-
plained, to a large extent, by language-independent 'Gricean' principles.
2. English nominal tautologies: semantic representations
My main claim concerning English nominal tautologies is this: English
has not one, but many, productive tautological patterns conforming to
the formula (ART) N
j
be (ART) Nj" If we want to state the meaning of
these patterns accurately, we must recognise this plurality of types, and
state the meaning of each one separately. In some cases, formal clues
enable us to separate the different types. In other cases, the form of
two different tautological patterns is the same, yet they cannot be
collapsed under a simple semantic representation because their mean-
ings differ in ways which cannot be accounted for in terms of context
or lexical differences. For this reason, some tautological sentences are
ambiguous. Thus A mother is a mother can mean either that a mother
can always be expected to act in a motherly way (even if she seems
different from other mothers), or else that one has obligations toward
one's mother. Similarly, A steak is a steak can mean either that there
isn't much difference in value between one steak and another (one is
neither much better nor much worse than another), or else that all
steaks are undeniably and reliably things of high value.
From a formal point of view, the following tautological constructions
(among others) can be distinguished in English:
Nabstr is Nabstr War is war; *Wars are wars, *Wars will be wars.
N
p1
are N
p1
Kids are kids; *The kids are the kids.
N
p1
will be N
p1
Boys will be boys; *A boy will be a boy.
A N is a N A party is a party; *The party is the party.
404 Boys will be boys: even 'truisms' are culture-specific
The N is the N The law is the law; *The war is the war.
N
l
is N
l
(and N
2
is N
2
) East is East, and West is West.
From a semantic point of view, however, the list of English nominal
tautological patterns (which I will call sub-constructions) would present
a different picture, as shown by the survey of sub-constructions offered
below. While this survey is not presented as an exhaustive one, it
covers most of the commonly used types.
I should add here that, although it seems appropriate and fruitful
to speak of English tautological sentences in terms of a set of sub-
constructions, we should also note that some such sentences have the
status of more or less set phrases, for example Fair is fair; Enough is
enough; A deal is a deal; Business is business. However, it is important
to recognise that they are not idioms, but rather particularly frequent
tokens of productive tautological patterns. Even the sentence Boys will
be boys, which must be regarded as the focal member of the class to
which it belongs, is not really idiomatic, since its meaning is strictly
parallel to that of other sentences based on the same pattern, for
example Students will be students or Teenagers will be teenagers.
Even if such sentences are seen as variations on the theme of Boys will
be boys, one cannot indulge in such substitutions with true idioms,
such as He kicked the bucket.
2.1. 'Realism' in human affairs
A 'realistic' attitude toward complex human activities is expressed by
the following syntactic formula:
Nabstr is Nabstr
Examples:
War is war. Politics is politics. Business is business.
*Wind is wind. *Sneezing is sneezing. *Wars are wars.
This sub-construction seems to be restricted to complex human activi-
ties, and apparently to those which involve human interaction. This is
perhaps linked to the complex character of the activity - which can be
seen as a special 'way of life', or as a world apart - and to 'inevitable'
negative aspects of this activity, which must be understood and tolerated.
English nominal tautologies: semantic representations 405
From a formal point of view, this sub-construction is marked by an
absence of articles (as well as by the singular number). One can say A
war is a war; but the meaning is different from that of War is war
(see 2.8 below).
The meaning encoded in the present sub-construction can be repre-
sented in the form of a number of interrelated components. These include
a reference to a supposed truism that complex activities of the specified
kind must have some undesirable consequences. The nature of these
consequences is viewed as well known, so that it would be superfluous
to spell them out. Further, there is a call for acceptance of those
undesirable consequences: since they are inevitable, there is no point
in getting oneself into a negative emotional state every time one
observes them. The 'nature' of the activity is such that it necessarily
entails the undesirable consequences in question. Hence, the War is war
type may be explicated as follows:
War is war
(a) everyone knows: when people do things of this kind
something bad can happen to other people because of this
(b) I know: someone can think:
this is bad
it should not be like this
(c) I think: one should not think this
(d) one should know:
it is always the same (when people do things of this kind)
it cannot be not like this
(e) I don't want to think: this is bad
(f) I don't want to feel something bad because of this
2.2. Tolerance for human nature
A tolerant, and also 'realistic', attitude toward human nature is expressed
by the following syntactic formula:
N are N
hum.pI hum.pI.
Examples:
Boys are boys. Kids are kids. Women are women.
Children are children. They are there to be put up with. (Pascal
1981:34).
406 Boys will be boys: even 'truisms' are culture-specific
The meaning of this sub-construction is of course very similar to the
preceding one, and one might well be tempted to assign it the same
semantic representation. But a semantic formula obliges one to be
explicit, and this in tum forces one to pay attention to detail and to
subtleties. When this is done, certain clear differences seem to emerge.
First, the component 'bad' seems applicable to the 'human activity'
type, but not to the 'human nature' type. Wars are said to be 'horrible';
politics is a 'dirty business'; the world of business is 'ruthless'. But
when one says that Kids are kids, one doesn't wish to imply anything
truly 'bad' about children: they may be noisy, boisterous, unruly, tire-
some, but not 'bad'. Accordingly, I would differentiate the relevant
components of the semantic representations as follows:
(a) everyone knows: when people do things of this kind
something bad can happen to other people because of this
(a') everyone knows:
people of this kind do things like this
one would want them not to do things like this
Furthermore, I would posit for the 'tautologies of human nature' an
additional component, which highlights their benign, indulgent tone:
(g) I think: people of this kind are not bad
All the other components are essentially the same: the realisation that
other people may complain about the undesirable aspects of the
situations discussed, the conviction that one should not do that, the
recognition of the predictability and the inevitability of those undesir-
able aspects, and the call for calm and acceptance. But given the
differences pointed out above, this call for acceptance will be inter-
preted differently in each case. In the case of complex human activities
such as war it will have the flavour of sober, wordly-wise resignation;
whereas in the case of the 'tautologies of human nature', it will sound
like a call for indulgence and tolerance. This is linked with the fact
that while in the case of human activity the undesirable aspects are seen
as a matter of (so to speak) grim necessity, in the human nature type
they are seen, rather, as a matter of human weakness.
13
Compare:
(d) it cannot be not like this
(d') they cannot be not like this
English nominal tautologies: semantic representations 407
This brings us to the following over-all formula for the Kids are
kids type:
Kids are kids
(a) everyone knows:
people of this kind do things like this
one would want them not to do things like this
(b) I know: someone can think:
this is bad
it should not be like this
(c) I think: one should not think this
(d) one should know:
all people of this kind are like this
they cannot be not like this
(e) I don't want to think: this is bad
(f) I don't want to feel something bad because of this
(g) I think: people of this kind are not bad
The indulgent and patronising attitude of human-nature tautologies is
particularly pronounced in the subtype with will (Boys will be boys),
where the 'immutability' of the phenomenon is presented as due to the
willful and uncontrollable spontaneity of a non-serious and pleasure-
seeking species. The following set of sub-components can be postulated
for this subtype:
(d) one should know:
all people (creatures?) of this kind are the same
they want to do things like this
because they want to feel something good
they will do them because of that
they cannot be not like this
The indulgent attitude encoded in this construction comes across
very clearly in the section entitled "Dogs will be dogs" in Christina
Johnson's (1978) book In praise of dogs in Australia. The section, which
is devoted to dog exercise, opens as follows: "Australia's extensive
beaches, ample parklands and wide open spaces offer ideal opportruni-
ties for dogs to revel in exuberant romps, wild gallops and invigorating
swims." It ends on the same note: "Frivolous though they sometimes
seem, these romps will not only keep the dog fit; they are essential to his
mental well-being." (Johnson 1978:32).
408 Boys will be boys: even 'truisms' are culture-specific
The expression "Dogs will be dogs" is of course jocular, but it is
instructive in so far as it highlights two important semantic aspects of
the category in question: 'unruliness' (mentioned also in Levinson's off-
the-cuff paraphrase) and fun-loving. The relevance of these two features
is also corroborated by some other types of English nominal tautologies,
to which we will turn next. But first, a formula for the Boys will be
boys type:
Boys will be boys
(a) everyone knows:
people of this kind do things like this
one would want them not to do things like this
(b) I know: someone can think:
this is bad
they should not do it
(c) I think: one should not think this
(d) one should know:
all people of this kind are the same
they want to do things like this
because they want to feel something good
they will do them because of that
they cannot not do them
(e) I don't want to think: this is bad
(f) I don't want to feel something bad because of this
(g) I think: people of this kind are not bad
2.3. Tolerance at 'special times'
Syntactic formula:
A N is a N
English seems to have a productive tautological pattern which ex-
presses a forebearing attitude towards pleasurable activities taking
place during 'special', privileged times, and on special, privileged
occasions, such as holidays, birthdays, parties and so on. One can say,
for example:
A picnic's a picnic. (Nesbitt 1980:222)
A party is a party.
A holiday is a holiday.
English nominal tautologies: semantic representations 409
A honeymoon is a honeymoon.
A game is a game.
One can hardly say, however:
?A Monday is a Monday.
?An autumn is an autumn.
?A morning is a morning.
(unless in a totally different sense, to be discussed below: 'one Monday
is not better or worse than another'). The privileges associated with
'special' times such as holidays, parties, etc., may entail inconveniences
for other people, but such inconveniences have to be excused and put
up with Gust as inconveniences due to the weakness of human nature
have to be put up with).
One cannot say, however: *A picnic will be a picnic, or *A holiday
will be a holiday, as one can say: Boys will be boys, because in the case
of picnics, holidays and such, the need for tolerance is based on the
'special' character of a certain time, not on a foreseeable character of
a 'species' of (willful) human beings.
A picnic is a picnic
(a) everyone knows:
at times of this kind people want to feel something good
because of this they want to do some things
they wouldn't do these things at other times
(b) I know: someone can think:
this is bad
it should not be like this
(c) I think: one should not think this
(d) one should know:
it is always like this (at times of this kind)
it cannot be not like this
(e) I don't want to think: this is bad
(f) I don't want to feel something bad because of this
(g) I think: this is not bad
Neither can one say: *Picnics are picnics, *Parties are parties,
*Honeymoons are honeymoons. Since the stress is here on the 'special'
rather than on the general and the predictable, the singular is apparently
more appropriate for the category in question; and it can in fact be seen
as an icon of this aspect of meaning. But since the 'special' character of
410 Boys will be boys: even 'truisms' are culture-specific
times, and situations, such as picnics, parties, or honeymoons, consists
in the suspension of normal 'rules' of life, and since this suspension
of rules is linked with a licence for pleasure-seeking activities, this
type is in fact closely related to the Boys will be boys type.
2.4. The limits of tolerance
The English saying:
Enough is enough.
may seem to express a sentiment radically different from that expressed
in sayings referring to human nature, such as Boys will be boys. The
latter is tolerant and indulgent, the former stern and implacable. In
fact, however, the two types are closer than they seem, as they share a
common theme: that of tolerance. In saying Enough is enough (or
similar utterances like A joke's a joke (but that's enough))the speaker
is not being stem and implacable in the sense in which he would be if
he said, for example, Stop it at once! Rather, he indicates that while he
recognises the need for tolerance and while he is quite prepared to be
indulgent for a time, now the limits of his tolerance have been reached.
Accordingly, the semantic representation of the 'stem and implacable'
sentence Enough is enough is in fact closely related to that of the indul-
gent sentences such as A picnic is a picnic or Boys will be boys:
Enough is enough
(a) everyone knows:
sometimes people want to feel something good
because of this they want to do some things
they wouldn't do these things at other times
(b) I know: someone (you?) was doing something like this for
some time
(c) I know: someone can think: this is not bad
(d) I think: one should not think this
(e) one should know:
one can do it for some time
one should not do it for a long time
(f) I want this person not to do it any more now
English nominal tautologies: semantic representations 411
2.5. Seeing through superficial differences
Syntactic formula:
A N
hum
is a N
hum
In the BBC television series 'Fawlty Towers' one of the characters
says at one point: A man is a man. The attitude to men conveyed in
this particular utterance is far from tolerant and indulgent. The speaker,
a woman, has just discovered a dead man in a cupboard, and feels
frightened; another woman tries to reassure her, pointing out that a dead
man can do her no harm; this, however, is countered with the somber
tautology quoted above.
It is interesting to note that the deep distrust towards men (even dead
men) has to be expressed in the singular (A man is a man, and not Men
are men), whereas patronising tolerance is always expressed in the plural
(Kids are kids, not A kid is a kid). The singular focusses on the serious
harm which is to be expected from an individual and which should not
be dismissed lightly. The plural refers to peccadilloes which have
already been perceived and which should be dismissed, because they are
trivial, and because they are not under the individual's control.
A number of eighteenth and nineteenth century collections of
English proverbs include one or another version of the following:
A man is a man, tho' he have but a hose upon his head
Stevenson (1949:1511) offers the following comment: "Some commen-
tators think that the meaning of this proverb is not clear, but it must
mean that a man's a man, whatever his attire, even if his poverty
compels him to wear a stocking for a cap."
Obviously, the attitude to men conveyed by the proverb is different
from that conveyed in the 'Fawlty Towers' quote. One says, in effect:
'don't trust any man'; the other says 'don't despise any man'. But it
is not difficult to find a common thread linking the communicative
import of the proverb and of the homophonous sentence used in
'Fawlty Towers'. Both imply that 'despite differences in appearance,
one man is not significantly different from others; in essence, all men
are the same'.
An examination of other English tautologies encoding a similar
message ('despite appearances, all Xs are the same') shows that in
different collocations and in different contexts this message can be
combined with a number of different attitudes. For example:
412 Boys will be boys: even 'truisms' are culture-specific
A man is a man (and not to be trusted, or, not to be despised).
A priest is a priest (and can be trusted).
A mother is a mother (and can be expected to care).
A lawyer is a lawyer (and can be relied upon as an expert).
A woman is a woman (and can be expected to show, at some
stage, a 'feminine weakness', even if she seems generally
tough).
This list of attitudes and expectations is of course far from exhaustive,
and it is doubtful whether an exhaustive list of this kind could ever be
compiled. It seems sounder, therefore, to assume that the different
attitudes and expectations involved are not part of the meaning of the
construction, and to look for a general formula, which would include
only what all tautologies of this kind have in common. Clearly, what
is needed is a formula which would fit the whole range of possible
expectations, and which nonetheless wouldn't over-predict (cf.
Boguslawski 1981a:l02).
Consider, for example, a situation where a girl giggles in a 'typically
girlish' fashion. The comment: (Ah, well), girls are girls would imply
that this is precisely what one could expect of any member of the
category 'girl'. On the other hand, the comment: (Ah, well), a girl is a
girl, if acceptable at all, would be seen as implying that one would not
have expected such 'girlish' behaviour of this particular girl, although
one would expect it of girls generally; and the comment would be taken
as a vindication of the general validity of the stereotype.
On the other hand, if a usually tomboyish girl exhibits some behaviour
regarded as typically girlish, one could say A girl is a girl. The differ-
ence between the singular and the plural pattern is, then, this: while both
patterns endorse a certain stereotype, the 'singular' one acknowledges
that in some particular case the appearances seem to be to the contrary -
and yet reaffirms the stereotype despite the appearances. By contrast,
the 'plural' one simply reaffirms the stereotype, without any references
to deceptive appearances. In addition, the 'plural' type implies, as we
have seen, benign tolerance for human weakness. But there is no sugges-
tion of benign tolerance in the singular type.
This brings us to the following semantic formula for the A man is a
man (or A girl is agirl) type:
A man is a man
(a) everyone knows: one can say: people of this kind are like this
English nominal tautologies: semantic representations 413
(b) I know: someone can think:
this person is not like other people of this kind
(c) I think: one should not think this
(d) one should know:
all people of this kind are the same
a person of this kind can be a little different from other
people of this kind
they cannot be very different
2.6. Recognising an irreducible difference
Syntactic formula:
N
1
is N
1
(and N
2
is N
2
)
As mentioned earlier, sentences such as:
East is East (and West is West).
White is white (and black is black).
refer to what the speaker presents as an irreducible difference, which
follows from the unique nature of each member of the set, and which
has to be accepted as irreducible. As a rule, the sets in question are
binary, and the two members can be seen as opposites. Formally, tautolo-
gies of this kind are distinguished by their complex character: they
are not merely tautologies but double tautologies; and although the
second part can be omitted, its presence is implied by the first part.
Consider also a parallel Polish example:
- Co pan wlasciwie rozumie, panie Stefanie, pod srowem
borowka? pyta pan Mareczek.
- Borowka - odpowiada wiolonczelista - to czarna Jagoda. A dla
pana to co?
- Borowka to borowka a czarna Jagoda to czarna Jagoda.
(Rymkiewicz 1984:9)
'''What exactly do you mean by the word borowka, Stefan?"
Mareczek asked.
"Borowka," the violinist replied, "means (a) blueberry. And for
you, what does it mean?"
"(A) borowka is (a) borowka, and (a) blueberry is (a) blueberry.'"
414 Boys will be boys: even 'truisms' are culture-specific
This sub-construction can be explicated as follows:
East is East (and West is West)
(a) everyone knows: one thing/person is not the same as another
(b) I know: someone can think:
if one can say something about X one can say it about Y
(c) I think: one should not think this
(d) one should know:
X is not like Y
one thing/person can be like another
one thing/person cannot be the same as another
It might be added that a similar (though perhaps not identical) mes-
sage can be conveyed in sentences referring to individual people. For
example, in the ABC television production 'Bewitched' (30.5.1985) one
of the heroes, Larry, suspects that two young women called Samantha
and Serena are in fact the same person, and he is glad to discover, while
speaking to one of them, that
You are you and she is she.
In the same program, somebody else asserts that
Samantha is Samantha and Serena is Serena.
In a similar vein, in C. S. Lewis' Screwtape letters the senior devil
makes the following assertion:
The whole philosophy of Hell rests on recognition of the axiom that one
thing is not another thing, and, especially, that one self is not another self.
My good is my good and your good is yours. (Lewis 1946:92).
The question of the relation between sentences of this kind and those
referring to finite (binary?) sets, such as East is East, is an interesting
one, but I will not pursue it here any further.
2.7. Tautologies of value
English has a number of tautological constructions focussing on the value
of certain things (and also people). I will discuss here four such sub-
constructions.
English nominal tautologies: semantic representations 415
The first construction emphasises the low value of certain categories
of things (and people). As an example, consider Kipling's sentence (from
'The Betrothed') quoted in Bartlett (1980):
A woman is only a woman, but a good cigar is a smoke.
At first, one might think that the meaning of the expression A woman is
only a woman is no different from the meaning of the expression A
woman is a woman, except for the additional semantic content of only.
But in fact, the variant with a 'belittling', minimising particle such as
only lacks what appears to be a crucial feature of the variant without
such a particle: the idea that 'despite appearances, this particular X is
not different from other Xs'. Thus in Kipling's sentence the stress is
not on 'some particular woman being no different from other women';
rather, it is on the low value of the entire class of women (as compared
with cigars).
Consider also the 'belittling' intention of the tautologies in the popu-
lar song from the film Casablanca:
You must remember this:
A kiss is just a kiss,
A smile is just a smile ...
Without some particle such as just or only, the same tautologies would
imply a different, and, in a sense, opposite attitude. Sentences such as
A kiss is a kiss, A smile is a smile would normally be interpreted as
implying undeniable commitment, value or significance. But the variant
with just asserts that all kisses (or smiles) are fairly insignificant, not
that this particular kiss (or smile) is not different in value from others.
Thus, in the semantic representation of tautologies with minimising
particles, the component 'I know: someone can think: this X is not like
other Xs of this kind' would not be appropriate.
Consider in tum the sentence: A point is a point, which I heard uttered
at an examiners' meeting at the AND (on 27 June, 1985). The speaker's
intention was, clearly, to imply that 'every point [for a semester unit
completed by a student] counts, and the importance of a single point
should not be underrated'. A similar sentiment is conveyed in the
following sentence:
'Must find it,' Fatty was saying. 'A shilling is a shilling.'
(Blyton 1980:57)
416 Boys will be boys: even 'truisms' are culture-specific
Here, too, the speakers do not seem to be concerned with an apparent
difference between one particular point and all other points, or
between one particular shilling and all the other shillings. Rather, they
are concerned with the value of any single point (or shilling); and they
insist that this value - though small - is nonetheless not negligible.
It should be noted that one could hardly use a plural in this sense:
?Points are points, ?Shillings are shillings. On the other hand, a plural
can probably be used in a related sub-construction, which is used to
remind people of the undeniable value of things which are not being
despised but whose value could nonetheless be occasionally forgotten.
Thus, one can say:
A steak is a steak.
A Mercedes is a Mercedes.
Money is money.
Eggs are eggs.
Some tautological utterances can be ambiguous in this respect. For
example, in saying: An egg is an egg, the speaker may mean either
'every small thing counts' or 'don't forget the well-known value
of eggs'.
Another related sub-construction requires an adverbial modifier of
time or space and allows the form Ns were Ns. For example:
In those days, women were women.
In Japan, women are still women.
In those days, children were children: they played, they studied,
and they did as they were told.
The adverbial modifier implies a contrast, and the tautological construc-
tion conveys approval, not indulgence.
Again, one must note that the 'well-known value' tautologies don't
seem to be necessarily concerned with differences (however superficial)
between one thing and another (of the same kind). For example, in
saying: A steak is a steak, the speaker doesn't seem to be claiming
that 'despite appearances, this particular steak is no different from
others'. Rather, he is emphasising the undeniable value of all steaks.
Finally, English has a tautology which expresses indifference to
individual differences, and which stresses the interchangeability and
the equal value of things (within a kind). Consider, for example, the
following exchanges:
English nominal tautologies: semantic representations 417
- How was the party?
- Oh - a party is a party.
- Do you want Nescafe or Maxwell House?
- It doesn't matter. Coffee is coffee.
The stress on the identical value of all Xs conveyed in sentences of
this kind makes them seem extremely close to the pattern A N
hum
is a
N
hum
And yet I believe that there is a difference between the two
patterns. Speakers who utter such tautologies sound blase, wordly-
wise, somewhat cynical. They want to be seen as people who don't get
unnecessarily enthusiastic or over-excited; they 'know' that usually
things are neither very good nor very bad; for example, A party is a
party: nothing to get excited about. To give a somewhat jocular illustra-
tion: a recent issue of Australian Natural History (The Australian
Museum, 1989, 22:540) reported new evidence on animals' responsive-
ness to individual faces. When presented with a series of slides, "the
sheep responded particularly to pictures of species with horns (an impor-
tant signal of sex and status), and the bigger the horns the more excited
they became". The picture accompanying the article bears the caption:
"A sheep is a sheep ... but not to a sheep." The message is clear:
individual sheep do differ in perceived value (to other sheep) - contrary
to the conventional wisdom embodied in the tautology.
In essence, the difference is this: one type of tautology refers to
'nature' and the other, to 'value'. For example, in saying:
A doctor is a doctor.
a speaker may be either referring to the supposed 'common nature' of all
doctors or to their 'equal value'. The intended message may be either:
(1) this particular doctor may seem different from others, but this is
deceptive, at the level of essential characteristics all doctors are the
same, and sooner or later their common nature will show; or (2) one
particular doctor may seem better (or worse) than others, but in fact
they are all the same: they do have some value, but one should not put
too much faith in them, and one should not attach too much importance
to the choice; this one is as good as any other. A third possibility is the
'undeniable value' one: all doctors are highly qualified professionals
and can be relied upon.
Naturally, ambiguities of this kind arise mainly with human nouns,
since it is mainly in the case of people (or animals) that one would want
to appeal to their nature. A sentence such as: A party is a party is
418 Boys will be boys: even 'truisms' are culture-specific
ambiguous, too, but here the choice is not between 'value' and 'nature',
but between 'value' and 'special time'; compare:
- How was the party?
- [shrugs] A party is a party.
It's noisy, but (after all) a party is a party!
Below, I will sketch in tentative semantic representations of the four
tautologies of value which have been singled out here.
A point is a point
(a) everyone knows: things of this kind are good
(b) I know: someone can think:
this thing is not like other things of this kind
it is something very small
(c) I think: one should not think this
(d) one should know:
all things of this kind are the same
they are all good
a thing of this kind can be small
it cannot be not good
A kiss is just a kiss
(a) everyone knows: things of this kind are like something small
(b) I know: someone can think:
this thing is not like other things of this kind
it is something very good
(c) I think: one should not think this
(d) one should know:
all things of this kind are the same
they are like something small
one thing of this kind can be a little less like something
small than another
it cannot be not like something small
A party is a party; Coffee is coffee; A girl is a girl
(a) everyone knows:
things of this kind are not very good, not very bad
(b) I know: someone can think:
this thing is not like other things of this kind
it is very good, or very bad
(c) I think: one should not think this
English nominal tautologies: semantic representations 419
(d) one should know:
all things of this kind are the same
one thing of this kind can be a little better or a little worse
than another
it cannot be very good or very bad
A steak is a steak; A Mercedes is a Mercedes
(a) everyone knows: things of this kind are good
(b) I know: someone can think: this one is not good
(c) I think: one should not think this
(d) one should know:
all things of this kind are the same
they are all good
one thing of this kind can be a little better or a little worse
than another
it cannot be not good
2.8. Tautologies of obligation
This sub-construction is expressed by the syntactic formula:
(ART) N is (ART) N
One use of tautological constructions, particularly widespread in
different languages, is concerned with obligations, and, more broadly,
with rules of human behaviour. Generally speaking, if a noun embodies a
modal meaning such as 'One should do X', then the pattern (ART) N is
(ART) N implies that the obligations in question must be fulfilled, even
if one prefers not to do so. Thus one can say The law is the law, as
the Australian judge Michael Kirby said recently ,on the ABC radio
current affairs program 'PM'.
Such sentences are extremely common in English. To quote a few
characteristic examples which I have recently heard or read:
A rule is a rule.
A bet is a bet.
A promise is a promise. (Doyle 1981:417,51.3)
A deal is a deal.
A test is a test.
An agreement is an agreement. (Doyle 1981:141)
420 Boys will be boys: even 'truisms' are culture-specific
Usually, this type can be recognised by the presence of a noun referring
to contractual undertakings such as bet, promise, deal, law etc. But
one can also use human relational terms such as father to refer to
generally accepted role obligations. Thus one can say A father is a
father, meaning that one has certain obligations toward one's father
which should be fulfilled, without any implication that this father may
seem to be different from other fathers. The speaker may mean that in
this particular case there seem to be some extenuating circumstances
(such as, for example, one's illness, or one's recent marriage) but that
the obligation 'remains an obligation' and has to be carried out. As
mentioned earlier, somebody might even want to say A war is a war,
meaning that one must carry out one's duty with respect to a war. The
two tautological expressions A war is a war and War is war differ in
their 'implicatures', and are not interchangeable.
However, when the meaning of a noun clearly implies an obligation,
then mass nouns without an article can be used in the sense under
discussion; for example Duty is duty. Such sentences don't imply that
promises, bets, the law, or duty 'cause bad things to happen', like wars
or politics (cf. 2.1 above); rather, they imply that certain rules of
human behaviour require compliance, regardless of their unpleasant or
inconvenient consequences.
It is also interesting to note that tautologies of obligation normally
require a singular form of the noun. One can hardly say ?Bets are bets
(cf. A bet is a bet), or ?Deals are deals (cf. A deal is a deal). A sentence
like Promises are promises seems to imply that 'Promises are no more
than promises, and can't always be relied on' - which is very different
from what we imply by A promise is a promise. It seems that plural
nouns can be used in the relevant sense only if they refer to items which
normally occur in sets, rather than singly: Rules are rules; Regulations
are regulations. This is parallel to the extension of the singular patterns
to fit the other kinds of tautologies which normally preclude the plural:
A girl is a girl. (absolute generalisation - or indifference)
Girls are girls. (tolerance for human nature)
*Girls are girls. (impossible as an absolute generalisation)
Twins are twins. (possible as an absolute generalisation)
A party is a party. (indifference - or absolute generalisation)
*Parties are parties. (impossible as an expression of indiffer-
ence)
Beans are beans. (possible as an expression of indifference)
English nominal tautologies: semantic representations 421
However, the ability to co-occur with the definite article (or with a
definite possessive) does seem to distinguish 'tautologies of obligation'
from the other tautologies which normally require the singular. Thus
Neighbours are neighbours can be used as an expression of indifference,
in response to the question Do you like your new neighbours? But the
sentence My neighbours are my neighbours conveys some sense of
obligation (perhaps, one must help one's neighbours, or one must be
loyal to one's neighbours).
Compare also the following sentences and their possible inter-
pretations:
a. Fathers are fathers.
b. Your father is your father.
c. A husband is a husband.
Example (a) has only one possible interpretation: tolerance for human
nature. Example (b) also seems to have only one interpretation: obliga-
tion (one must fulfil one's obligations toward one's father). Example (c)
has as many as four interpretations: obligation (one must fulfil one's
obligations toward one's husband), appreciation (everyone knows that
there is something good about having a husband), indifference (one
husband is neither better nor worse than another), and absolute gener-
alisation (all husbands are essentially the same; one knows what to
expect from them).
I conclude that, for both semantic and syntactic reasons, tautologies
of obligation cannot be subsumed under any other type, and must be
recognised as a separate tautological sub-construction. The meaning
encoded in such sentences can be spelt out as follows:
The law is the law; A promise is a promise; A father is a father
(a) everyone knows: all people have to do things of this kind
(b) I know: someone can think:
this time this person doesn't have to do this thing
(c) I think: one should not think this
(d) one should know:
all things of this kind are the same
one thing of this kind can be a little different from other
things of this kind
one cannot not do this thing because of this
It is interesting to note the links between 'tautologies of obligation'
such as The law is the law and 'tautologies of grim necessity' such as
422 Boys will be boys: even 'truisms' are culture-specific
War is war, links which could be summarised in the two key notions:
'cannot' and 'don't want'. For example, wars have aspects that people
'don't want'; and these aspects 'cannot' be avoided. Similarly, the law
may force people to do things that they 'don't want' to do; and yet
they 'cannot' not do them. It would be easy to phrase the explications
of the two types in identical terms, and thus to reduce the two types to
one. I think, however, that this would be a false economy. The minimal
pair mentioned earlier:
War is war. ('realism, grim necessity')
A war is a war. ('obligation')
shows that in fact the two types are truly distinct in English (although
they would not necessarily be distinguished in some other language).
I would add, however, that although the tautologies of obligation
occur in many different languages, and can be presumed to express an
attitude which is largely language-independent, in some languages it is
counterbalanced by the existence of constructions whose communicative
import is quite different. For example, in Polish one can say not only:
Prawo jest prawem.
'The law is the law.'
as one does in English, but also:
Prawo prawem, ale trzeba jakos iyc.
'The law [is] the law, but one has to live somehow.'
thus excusing lawlessness rather than endorsing the validity of laws. A
sentence of this kind indicates that while the speaker accepts the immu-
table validity of laws (or whatever rules are referred to) theoretically,
in practice he expects that people will find some ways around those
rules, and that this, too, should be understood and excused. This use
of tautological expressions, which indicates a dismissal of certain
phenomena rather than acceptance, seems to be alien to English. But
in Polish, tautological expressions are often used in such a dismissing
way. For example, the Polish expression:
tarty iartami.
'Jokes (NOM.PL) [are] jokes (INSTR.PL).'
translates into English as 'jokes aside' (as does its Russian equivalent
Sutki sutkami) and not as A joke is a joke. It indicates neither indulgence
nor impatience but simply a lighthearted dismissal ('jokes are fine, but
Some comparisons from Chinese and Japanese 423
let's now speak seriously'). Similarly, an expression such as Prawo
prawem ('the law is the law, but ... ') doesn't challenge the validity of
laws; it merely 'sets them aside' ('laws are all very well, but ... ').
3. Some comparisons from Chinese and Japanese
3.1. Chinese concessive tautologies
Chinese has some tautological patterns which seem more or less similar
to the English ones; but it also has patterns which are distinctly different.
(My discussion of Chinese tautologies is based on data and analysis
provided by Luo 1988.) One type which has no equivalent in English
is what one might call 'concessive' tautologies. They can be illustrated
with the following examples:
Qtnqi dao shi qfnqi, (jiushi bu tai qfn).
relatives PRT are relatives (but not very close)
'Relatives though they are, they are not very close.'
Jingzi dao shi jingzi (jiushi zhaoren bu qfng chu).
mirror PRT is mirror (but person not clear see)
'Though it is a mirror, it does not reflect the image clearly.'
Nanren shi nanren, (ki queshiio nanren wei).
man is man (but lacks manhood)
'Although he is a man, he lacks manliness.'
As these examples show, the 'concessive' tautology has the form of a
subordinate clause, which precedes the main clause. The subordinate
clause states an 'undeniable truth' but the main clause contradicts this
truth with respect to a specific instance: since this particular entity (X)
belongs to a certain kind (X), one might expect that it will have certain
properties, generally seen as characteristic of that kind; and yet, the
speaker points out, this particular X (X) doesn't have the properties
in question.
Tentatively, I would represent the meaning of this type of tautology
(as described by Luo) along the following lines:
424 Boys will be boys: even 'truisms' are culture-specific
Nanren shi nanren
(a) everyone knows: one can say:
people/things of this kind are all the same
(b) I know: because of this someone can think: this one is like
this
(c) I think: one should not think this
(d) one should know:
a thing of this kind can be not like this
this one is not like this
The function of this pattern can be seen as the opposite of that of the
common tautologies of the English language. In English, most nominal
tautologies emphasise the sameness of the members of a given category;
they acknowledge that one member may differ from another in some
superficial characteristics but they stress the commonality of the essen-
tial features. By contrast, the Chinese concessive tautology acknowl-
edges the sameness of some characteristics (presumably, the definitional
ones), but emphasises individual differences; it can be seen, therefore,
as a warning against absolute generalisations. If the English tautologies
of several kinds defend and reinforce the stereotype (A man is a man,
Boys will be boys, A Mercedes is a Mercedes), the Chinese concessive
tautologies warn against an excessive confidence in stereotypes.
This is not to say that Chinese doesn't allow the use of tautologies
for making absolute generalisations and for underscoring stereotypes.
It does, as in the following sentences (which usually include a modal
adverb):
Nuren z6ng shi nuren.
'Women are (always) women.'
Miigu6ren daodi shi meigu6ren.
'Americans are Americans, (after all).'
Qinqi zhongjiu shi qinqi.
'Relatives are relatives (after all).'
Lang z6ng shi lang
'Wolves are (always) wolves.'
Interestingly, the range of possible interpretations for such generalisa-
tions appears to be even wider in Chinese than it is in English, as they
can apply not only to 'human weakness' or to 'human/animal nature',
Some comparisons from Chinese and Japanese 425
but also to posItIve aspects of human relations. Thus, it seems less
natural to say in English than in Chinese:
?Relatives are relatives.
?Friends are friends.
?A relative is a relative
?A friend is a friend.
meaning that these relationships are invariably good and invariably
reliable, although the exact range of use of such tautologies requires
further investigation.
But while the English 'generalising' tautologies do have some counter-
parts in Chinese, the Chinese concessive tautologies do not seem to
have any counterparts in English. On the contrary, all English nominal
tautologies seem to have a 'final', 'topic-closing quality' about them,
epitomised in the common English tautological expression That's that.
By contrast, nominal tautologies in Chinese - and, I might add, in
many other languages of the world - do not seem to have such a
'topic-closing' quality. This is why they can often be used as a premise
in a concessive or quasi-concessive sentence of one kind or another, and
it is very interesting to see what kinds of 'undeniable truths' can be
undermined in a given language (and culture) in this way.
For example, it is interesting to compare Chinese concessive tautolo-
gies with Polish and Russian concessive tautologies mentioned earlier,
such as the following:
Prawo prawem, ale (Polish)
law:NOM.SG law:INSTR.SG, but .
'The law is the law, but ... '
Przyjain przyjaini{l, ale (Polish)
Druiba druzboj, no (Russian)
friendship:NOM.SG friendship:INSTR.SG, but .
'Friendship is all very well, but ... '
Polish and Russian tautologies of this kind can be said to be iconic:
the identity of the subject and the predicate nominal represents the
'undeniable truth' aspect of the sentence, whereas the absence of the
copula and the instrumental (rather than nominative) case of the predi-
cate nominal de-emphasises that identity and casts a doubt upon its
absolute validity. It should be pointed out, however, that Polish and
Russian concessive tautologies of this kind are narrower in scope than
the Chinese concessive tautologies illustrated earlier, being restricted to
abstract concepts such as 'joke', 'friendship', 'law', or 'promise'; unlike
426 Boys will be boys: even 'truisms' are culture-specific
the Chinese type, they cannot apply to concrete individual entities like
a person or a mirror. If the Chinese concessive tautologies warn
against absolute generalisations, the Polish and Russian ones call for
a flexibility of attitudes.
3.2. 'Irreducible difference', Chinese style
Like English (see 2.6 above), Chinese has nominal tautologies which
proclaim an 'irreducible difference' between two categories or two
entities. For example:
Dtng shi ding, mao shi mao.
'Ding is ding, mao is mao (one should keep ding from mao),'
(where ding represents 'a Heavenly stem' and mao, 'an Earthly
branch').
Shfsi shi shfsi, sishf shi sishf.
'Fourteen is fourteen, forty is forty.'
Tii shi ta, wo shi wo.
'He is he, I am I.'
However, as pointed out by Luo, the meaning of such tautologies is not
quite the same as that of their closest counterparts in English. In the
Chinese examples, unlike the English ones, the speaker combines the
statement of an 'irreducible difference' with a moral judgment: 'X and
Yare not the same; this is good; one shouldn't try to change it'. In
the English examples, the speaker points to an 'irreducible difference'
as to something that should be recognised, without implying, however,
that it is good and that it should be preserved.
The similarities and the differences between the two meanings (the
English and the Chinese ones) can be portrayed as follows (C for
Chinese, E for English):
Ding shi ding, mao shi mao (C) cf. East is East and West is
West (E)
(C/E) everyone knows: one thing/person is not the same as
another
(C/E) I know: someone can think:
if one can say something about X, one can say it
about Y
Some comparisons from Chinese and Japanese 427
(C/E) I think: one should not think this
(C/E) one should know:
X is not like Y
(E) one thing can be like another
(E) one thing cannot be the same as another
(C) this is good
(C) one should not think: X can be like Y
According to Luo (1988), Chinese allows also for multiple (rather
than binary) tautologies of 'irreducible difference'. He illustrates this
with a famous (copula-less) example which epitomises Confucius'
social doctrine:
Jun jiin, chen chen, fil fil, Zl Z1.
'A ruler is a ruler, a minister is a minister, a father is a father,
and a son is a son.'
Luo (1988) comments: "Written more than two millennia ago, this
doctrine was accepted as infallible law and has influenced Chinese
ideology ever since: what Confucius wants to express is that all the
social hierarchies must be kept distinct and should not be altered or
confused." Arthur Waley's rendering of Confucius' tenet (quoted by
Luo) supports Luo' s interpretation of this pattern:
Let the prince be a prince, the minister, a minister,
the father, a father, and the son, a son.
It supports also the components included in the explication of the Chi-
nese pattern (in contrast to the English one):
(X is not like Y)
this is good
one should not think: X can be like Y
3.3. Chinese tautologies of unreserved praise
The last type of Chinese tautologies to be mentioned here is very differ-
ent from anything that we have seen in English. It conveys praise and
admiration towards an entity whose all parts or all aspects are seen as
genuinely good. Typical examples are:
428 Boys will be boys: even 'truisms' are culture-specific
(Kan ren jia nei tiaojian:) caidian shi caidian, bingxiang shi
bingxiang, xiyiji shi xiyiji.
'(Just look at what they have:) colour TVs, fridges, washing
machines, (all kinds of things)!'
Tao shi tao, ZI shi Zi, xing shi xing, (guo yuan Zi shenme dou you).
'Peaches, plums, apricots - there are all kinds of fruit in the
orchard.'
Td yan zhenhao, yiinshen shi yiinshen, shenduan shi shenduan,
banxiang shi banxiang.
'S/he performs brilliantly - rich in expression, elegant in
posture and marvellous in appearance.'
I am not going to propose a full explication for this type of tautology, but
I would like to sketch, and to discuss briefly, some possible components.
First, there seems to be a component of 'total praise': 'one can say all
good things about this.' Possibly, there is an implication that 'everyone
knows' what the relevant aspects to be assessed would be for the cate-
gory in question. For example, a good orchard would be expected to
have peaches, plums, and apricots, all of good quality. Furthermore, all
the aspects which are enumerated are individually assessed against
an imaginary model, perhaps again with reference to common
knowledge (everyone knows what 'real' peaches should be like - these
peaches are like this; and so on). Finally, if Luo is right, this pattern
expresses also the speaker's admiration for the genuine quality of
the individual aspects of the entity in question and for the 'all-roundness'
of its perfection.
Some of the motives mentioned above (for example, 'everyone
knows') correspond to those which we have established for the nominal
tautologies of the English language; but some are clearly different.
It seems obvious that the exact semantic value of this last type of
Chinese tautology cannot be worked out on the basis of any 'natural
logic': it is language- and culture-specific.
Japanese has a number of nominal tautologies which correspond,
more or less, to some of the types distinguished here for English.
However, it has also some forms without any counterpart in English.
One such type ('what seems impossible is really possible') was already
mentioned in the Introduction. I will now discuss briefly two further
ones (on the basis of data and analysis provided by Field 1988).
Some comparisons from Chinese and Japanese 429
3.4. Japanese tautologies of 'a matter of course'
Tautologies of 'a matter of course' (Field's term) can be illustrated with
the following examples:
Keireki ga keireki dakara ...
personal history SBJ pers.hist. COP-BECAUSE
Toshi ga toshi dakara ...
age SBJ age COP-BECAUSE
Kao ga kao dakara ...
face SBJ face COP-BECAUSE
The general pattern is, then:
N. ga N. dakara ...
J J
As pointed out by Field, in tautologies of this kind the subject particle
ga cannot be replaced with the topic particle wa. This feature distin-
guishes the type in question from other types of nominal tautologies,
which normally take wa, not ga, for example:
Yakusoku wa yakusoku da.
rule TOP rule COP
'Rules are rules.' (and have to be obeyed)
Gaijin wa gaijin da.
foreigner TOP foreigner COP
'Foreigners are foreigners.' (implied: criticism and resignation)
Tautologies of 'a matter of course' imply that the reason of some state
of affairs is self-explanatory. They could be, therefore, compared with
the English expression: N being what it is ... For example:
Toshi ga toshi dakara ...
age SBJ age COP-BECAUSE
'His age being what it is ... '
(it's no wonder he's puffing by the time he reaches the fourth
floor)
I am not going to propose a full explication for this type of tautology.
One familiar component, however, is fairly easy to discern: 'everyone
knows'. Another familiar component which may come to mind is 'all Ns
are the same': for example, 'all old men are the same (in the relevant
430 Boys will be boys: even 'truisms' are cUlture-specific
respect)'. But this can hardly be applicable to strictly individual circum-
stances, such as somebody's 'personal history'. A better guess might be:
'it would be the same for anyone (of that age, with that kind of back-
ground, with such a face, and so on).' But the matter requires further
investigation.
3.5. Japanese tautologies of irrelevance
Tautologies of irrelevance can be illustrated with the following
sentences:
Hito
person
wa hito
TOP person
da.
COP
Otto wa
husband TOP
otto da.
husband COP
Field (1988:9) observes that Hito wa hito da could be "said by a teenager
to her mother who refuses to let her go to a rock concert because the
other mothers in the neighbourhood are doing the same", and it would
mean that the teenager is implying to her mother that she wants the
other mothers kept out of the argument; that other people are irrelevant
to the issue at hand.
Otto wa otto da "could be uttered in a situation where a group
of housewives are discussing whether or not they should still remain
friends with a Mrs. Tanaka whose husband has just been gaoled for
fraud. The housewife who utters the tautology implies that Mrs.
Tanaka's husband should be kept out of the issue - because the
husband is 'bad' does not mean that Mrs. Tanaka is bad too."
As a first approximation, the meaning of such tautologies can be
represented as follows:
Hito wa hito da
(a) everyone knows:
people are not the same
different people can do different things
(b) I know: someone can say:
if one can say something about one person
then one can say it about another
because of this, one should think about other people
Verbal tautologies 431
(c) I think: one should not think this
(d) I don't want to think now about other people
As this formula indicates, the Japanese tautologies of irrelevance are
closely related to the English tautologies of 'irreducible difference'. In
fact, sentences such as Hito wa hito da can be seen as functional equiva-
lents of English sentences like Samantha is Samantha and you are you.
But the English sentence emphasises not only the difference but also
the uniqueness of the entities in question (hence the common use of
proper names or personal pronouns in this type of tautology). By con-
trast, the Japanese tautologies of irrelevance do not refer to people's
unique features or unique situations; rather, they simply stress that
people are not the same, and more specifically, that different people
can do different things.
4. Verbal tautologies
The bulk of this chapter has been devoted to various kinds of nominal
tautologies, and limitations of space preclude a similarly detailed discus-
sion of what might be called verbal tautologies. However, before any
general discussion of a possible invariant or invariants of all tautologies
can be fruitfully undertaken, one needs to look briefly at at least
some verbal tautologies - from English, and not only from English.
This is the purpose of this section.
4.1. Future events
Tautologies referring to future events often have a rather fatalistic ring
about them. A somewhat fatalistic acceptance of future events (including
misfortunes) is encoded, for example, in the Spanish saying Que
sera sera ('what will be will be'), whose meaning can be explicated
as follows:
Que sera sera
I know: we don't know what will happen to us
I know: something bad can happen to me
I don't want to think about it
432 Boys will be boys: even 'truisms' are culture-specific
I don't want to feel something bad because of it
I know: I cannot think:
'if I do something these bad things will not happen to me'
The first of these components is reflected in the song:
Que sera sera,
Whatever will be will be,
The future's not ours to see,
Que sera sera.
Tautologies of this kind ('whatever will be will be ') seem to be
very common in different languages of the world. But here, again,
one should beware of hasty generalisations and one should not mistake
similarities for identities. For example, as pointed out earlier, the literal
Polish counterpart of the Spanish saying conveys also the idea of
personal risk, and an attitude of foolhardy bravery. The Polish saying
requires, therefore, some additional and different semantic components.
I propose the following:
Co ~ z i e to ~ z i e
I want to do something
I will do it because of this
I know: something bad can happen to me because of this
I don't want to think about it
The need for such different components highlights the culture-specific
twist which Polish culture gives to the conventional and (one might
think) universal wisdom. This is entirely consistent with the characteris-
tic Polish ethos, which has been described and documented elsewhere
on the basis of other kinds of evidence, both linguistic and non-linguistic
(see Wierzbicka, to appear, chaps. 2 and 5).
A different kind of tautology referring to future events is illustrated
by the English sentence mentioned by Levinson (1983: 111):
Either John will come or he won't.
Levinson's gloss, quoted earlier, reads: "Calm down, there's no point
in worrying about whether he's going to come because there's nothing
we can do about it." But though sound in essence, this gloss is clearly
ad-hoc and arbitrary in phrasing, which makes any comparison of this
particular pattern with other tautological patterns impossible. Using the
Verbal tautologies 433
controlled semantic metalanguage employed in the present work, we
can propose the following explication:
Either he will come or he won't
I know:
it can happen: he will come
it can happen: he will not come
I know:
someone can think: this is bad
someone can feel something bad because of this
I don't want to think about it
I don't want to feel something bad because of it
I know: one of these things will happen
I know: I cannot think: 'if I do something it will not happen'
This is not nearly as 'fatalistic' as the Spanish saying. To begin
with, the speaker is talking about future events which are probably
'undesirable' rather than 'bad', and which in any case are not presented
as directly affecting the speaker (or the addressee). There is no question
here, then, of 'something bad happening to me'. Furthermore, there is no
reference here to the (supposed) 'unknowability of the future in general'
(no 'we don't know what will happen to us' component). Finally, there
is some positive certainty, from which the speaker can take some
satisfaction, or at least some reassurance: 'I know: one of these
things will happen'. What is unknown and unpredictable to us is, above
all, other people's future actions; but these can, presumably, be planned
and controlled by the individuals involved.
All in all, the English pattern seems more compatible with the view
that human beings have some control over their future; not over other
people's actions, because this is their (other people's) domain of free-
dom; but there is some ground for calm in the thought that every
person has control over their own actions; and also, that to a logically-
minded person there are some certainties in life ('I know: one of these
things will happen'). This quiet certainty derived from a rational,
logical thinking ('either it will happen or it won't') can also be inter-
preted as a kind of control over the world. This is different from the
more passive, more indolent tone of Que sera sera.
434 Boys will be boys: even 'truisms' are culture-specific
4.2. The immutability of the past
Examples:
Let bygones be bygones.
Co bylo to by/o. (Polish)
'What has been has been.'
Tautologies expressing an acceptance of past events may seem to be a
mirror image of those expressing an acceptance of things to come. And
indeed, when one compares the English saying Let bygones be bygones
with the Spanish saying Que sera sera, one has to acknowledge a
number of close parallels. In both cases, the events referred to must be
seen as 'bad'; in both cases, they must be seen as unchangeable; and
in both cases the speaker refuses to think about the bad events in
question and to let them generate 'bad feelings'.
But the parallel is far from perfect, and the details of the interpretation
are by no means predictable. In particular, the sentence Let bygones be
bygones must refer to 'bad things' which have occurred between the
interlocutors, that is to say, to a past conflict between them. Nothing of
the kind is implied by tautologies referring to future events. Furthermore,
the formula Let bygones be bygones conveys an active call for peace.
The full meaning of Let bygones be bygones can perhaps be represented
as follows:
Let bygones be bygones
I know: we can think:
'some time before now this person did something bad to me'
we can feel something bad because of this
I don't want us to think about it any more
I don't want us to feel anything bad because of it any more
we know:
it happened before now
we cannot think:
'if we do something these bad things will not happen'
I think we should not think about it any more
In the speaker's view, the past is immutable. There is no point in
worrying about it because nothing can be done about it, anyway. It
cannot be thought that 'if we do something, these bad things will not
happen': they have already happened.
Verbal tautologies 435
In his call for peace, the speaker does not reiterate any accusations.
Presumably, it is the addressee who has done something bad to the
speaker (or so the speaker sees it); but the message is tactfully diffused
so as not to point a finger at the addressee; it is: 'we can think: someone
did something bad to me' rather than 'I know: you did something bad
to me'.
Again, it appears that exhortations of this kind ('let's forget the past')
are very common in different languages of the world; and coming
across yet another saying such as the following Chinese one (Luo
1988:12), provokes a feeling of deja vu:
Guoqu shi guoqu,
past is past
xianzai shi xianzai.
present is present
It must be noted once more, however, that superficially analogous
sayings in other languages may convey meanings different from that
encoded in the English one. For example, the Serbian saying: Sto je bilo
bilo je, 'What has been has been', can convey a nostalgic longing for
good events of the past which have irrevocably vanished. It is often
used with reference to a past love, as in the popular song:
Sto je bilo bilo je,
Vratiti se neee ...
'What has been has been,
It will not return ... '
Apparently, the same attitude can be conveyed by the Russian saying
eto bylo to bylo ('what has been [that] has been'), as its use in Bulat
Okudzava's poem Lunin v Zabajkal' e illustrates:
Neuzto eto bylo to by/o?
I gvardija vas pozabyla,
i daze ne snites' vy ej ... (Okudzava 1976:47)
'Is it possible that what has been has been?
And that the regiment has forgotten you,
and that they don't even have dreams about you any longer ... '
The literal English translation (What has been that has been) is
not only unidiomatic but is simply puzzling and hard to interpret to
native speakers. And the literal Italian translation is normally used in
yet another sense, illustrated by the popular love song:
436 Boys will be boys: even 'truisms' are culture-specific
Quel ch' estato estato,
Non importa s' efinita ...
'What has been has been,
It doesn't matter if it is finished ... '
In this use, the speaker affirms the immutable value of things past
('nothing can rob one of the treasure of one's past good experiences').
The equivalent Polish sentence, Co byio to byio, is normally used to
dismiss thoughts about some past events which were 'bad'. To use it
in a sense parallel to that of its Italian counterpart one would have to
mark it in a special way, preceding it with the particle conjunction
ale 'but':
Ale co byio to byio.
'But what has been that has been.' (i.e. 'but nothing can erase the
good things that have happened to us')
Being used in this way (with an initial ale) the sentence is an instance
of a more comprehensive Polish construction referring to past events,
which can be illustrated with the following sentences:
Ale cosmy ~ najedli tosmy ~ najedli.
'But what we have eaten, that we have eaten.'
Ale coscie widzieli toscie widzieli.
'But what you have seen, that you have seen.'
The construction in question (marked formally with the particle/
conjunction ale) is quite productive, its range of use being restricted
only by its meaning. Roughly speaking, this meaning can be character-
ised as follows: 'despite all the bad things that have happened to person
X, X should recognise and relish the gain which these events have also
caused; X should recognise that this gain has a lasting value, which
nothing can take away from X'. Typically, the events in question
concern the speaker and/or the addressee, but a third person use is not
totally excluded. What does seem to be required is a pronominal, or
rather a zero-anaphoric reference to the experiencer.
The meaning conveyed by this construction can be represented,
roughly, as follows:
Ale co S to S
I know:
some bad things happened to us
Verbal tautologies 437
we could feel something bad because of this
I think:
we should not think about it any more
we should not feel something bad because of it
we know: some good things happened to us at the same time
I think:
we should think about these good things
we should feel something good because of this
we know:
it happened before now
we cannot think: 'these good things will not happen to us'
Thus, the (perceived) immutability of the past can be invoked as a
justification or source of support for different emotional attitudes: for
resignation, for nostalgia, but also for a comforting certainty; for
generous peace-making, or for resolutely looking ahead, without
dwelling fruitlessly on past wrongs or past mistakes.
Attitudes of this kind can become culturally entrenched and linguisti-
cally encoded. When they are dressed up, in a particular language and
culture, as unquestionable 'universal wisdom', they can, one must
presume, become further entrenched, and may become an important
feature of a given cultural tradition.
As one last example of this kind let us consider the English tautology:
What's done is done (exemplified in a cartoon in AND postgraduate
newsletter Antitheses 1989, 1.4:1). This expression, as far as I know,
doesn't have any (tautological) counterparts in the language of Eastern
or Southern Europe. The meaning of this tautology can be represented
as follows:
What's done is done
I know:
someone did something bad
someone can think: this is bad
someone can feel something bad because of it
I think:
we should not think about it any more
we should not feel something bad because of it
we know:
it happened before now
we cannot think:
'if someone does something this bad thing will not happen'
438 Boys will be boys: even 'truisms' are culture-specific
It seems clear that this is yet another example of a pragmatic and
rationalistic folk philosophy which is associated with the English
language: the speaker takes stock of the situation, briskly acknowl-
edges some past mistakes, and resolutely looks ahead, trusting in logic,
experience, and action.
I suspect that while the use of tautological constructions with respect
to past and future events shows striking cross-linguistic similarity,
languages differ more in the way they use such constructions with
respect to the present. Consider, for example, the awkwardness of the
English translations of the Italian tautologies in the following exchange
from Manzoni's novel 'The betrothed':
'Tonio! non mi riconosci?'
'A chi la tocca, la tocca', rispose Tonio ...
'L' hai addosso eh? povero Tonio; ma non mi riconosci piu? '
'A chi la tocca la tocca', replieD quello ... (Manzoni 1972:794).
Bohn's translation of this passage reads as follows:
'Tonio! don't you know me?'
'Whoever has got it, has got it [the plague] ... '
'It's on you, eh? poor Tonio: but don't you know me again?'
'Whoever has got it has got it', replied he ... (Manzoni 1914,
2:615).
In Colqhoun's translation, the same passage reads:
'Tonio! Don't you recognise me?
'What comes to one, comes to one', replied Tonio ...
'You've got it [the plague] on you, eh?' Poor Tonio; but don't
you recognise me any more?'
'What comes to one, comes to one,' replied the other ....
(Manzoni 1968:457).
It appears that the only pattern which can be used in English to express
acceptance of the present is the pseudo-conditional (If S, S) one:
If you've got it you've got it.
If you have freckles you have freckles.
If it rains it rains.
Is there a semantic invariant? 439
5. Is there a semantic invariant?
I have argued that the English tautological expressions don't all have the
same inherent value, and that their different 'implications' cannot be
fully attributed to context. It goes without saying that the similarities
between different tautological constructions are no less real. It is much
easier, however, to assert that such similarities exist, or to hint at their
nature by means of vague or semi-metaphorical cover terms, than to
establish with some rigour what the common components of English
'tautologies' might be.
Levinson mentions "a dismissive and topic-closing quality" as a
shared feature of tautological constructions of different kinds. To make
these hints clearer, let's assume that the 'topic-closing quality' can be
spelt out in the form of the following semantic component:
I don't want to talk about it any more.
One could argue with some plausibility that a component of this kind
is indeed present in the meaning of some tautological expressions, such
as, for example, Enough is enough and That's that. A wish for no more
to be said on the topic seems associated with the tautologies in the
following Italian examples, and their English translations:
"Ma se non ne voglio saper nulla di quelle cose", diceva.
"Quante volte 10 devo ripetere, che quel che e andato e an-
dato?" (Manzoni 1972:735)
"But what if I don't want to hear anything more about these
things?'" he kept on saying. "How many times must I go on
repeating that what's gone has gone?" (Manzoni 1968:416)
"Chetatevi un po' ", disse don Abbondio; "che gia Ie chiachiere
non servono a nulla. Quel ch' efatto efatto: ci siamo, bisogna
starci." (Manzoni 1972:723)
"Keep quiet, will you, for a bit?" said Don Abbondio.
"Chattering's no use now. What's done is done. Here we are,
and here we must stay. It's all in the hands of Providence ... "
(Manzoni 1968:407)
But it seems doubtful that all tautological expressions can be interpreted
along those lines. For example, when the woman in 'Fawlty Towers'
points out that A man is a man, she is indeed 'dismissing' her friend's
440 Boys will be boys: even 'truisms' are culture-specific
attempt to reassure her, but is she also conveying the idea 'I don't want
to talk about it any more' ?
And even the expression which may seem to be 'topic-closing' par
excellence, That's that, doesn't always seem to be interpretable as
'topic-closing'. For example, in the BBC television program 'All
Creatures Great and Small', a veterinarian examines a little girl's dog
which has been run over by a car, and then says gently to the little
girl: I'm afraid that's that. Certainly, there is a 'finality' about the
vet's verdict, but can his attitude be described as 'I don't want to talk
about it any more' ? Presumably not.
Finally, when we say Boys will be boys do we really want to say 'I
don't want to talk about it any more'? Consider the following example:
"Just look at them swilling all that beer, and telling dirty jokes!
It's enough to make you sick."
"Oh, I don't know, you must let them off the leash a bit now and
then and remember that boys will be boys, poor things." (Cowie
1976, 2:75)
It is not at all clear that the second speaker is trying to close the subject.
The word 'dismissive' fits hislher attitude better, but again, what exactly
does this label mean? If we interpret it in the sense 'I don't want to
worry about it' or 'I think we shouldn't worry about it', then these
interpretive formulae would indeed seem to fit several other tautological
expressions. But would they fit, for example, sayings such as The law
is the law? One can of course say that this saying, too, is 'dismissive',
because here the speaker is dismissing somebody's objections to a
particular law, or somebody's attempts to get around a law. But
surely, in this case, 'dismissive' does not mean 'I don't want to talk
about it any more'; and if it doesn't mean that, then what does it mean?
The same question would apply to another idea which is sometimes
mentioned in connection with tautological constructions - the idea of
'acceptance'. Speaking loosely, one can easily agree that by saying War
is war or Kids are kids or Whatever'll be will be, the speaker 'accepts'
the situation. But in what sense does he accept it?
For a number of tautological expressions, I have proposed the compo-
nent 'I don't want to feel something bad because of this'. But when the
vague term 'acceptance' is replaced with a clear semantic formula which
can be checked against the actual use of different expressions, it
becomes clear that the formula in question is not equally appropriate
in all cases. For example, when someone says A snake is a snake, can
Is there a semantic invariant? 441
this be reasonably interpreted as conveying 'acceptance' in the sense of
'one shouldn't feel something bad because of that' ? Presumably not.
In fact, it would seem that the component of 'acceptance' in the sense
just explained (roughly, 'there is no point in getting upset about it') is
never present in nominal tautologies with a singular noun and indefinite
article. For example, despite all the differences in their respective inter-
pretations, none of the tautologies below seems to imply acceptance:
A Mercedes is a Mercedes (one should know the value of
Mercedes cars)
A point is a point (one shouldn't undervalue it just because
it's small)
A teacher is a teacher (one should respect teachers)
A kiss is just a kiss (one shouldn't overvalue a kiss)
A promise is a promise (one shouldn't break a promise)
A party is a party (one shouldn't overestimate the differences
between one party and another)
On the other hand, acceptance is not tied to a single formal pattern, since
it is compatible with plural nouns (without an article): Boys will be boys;
with singular nouns referring to complex activities and unaccompanied
by an article: War is war; with singular nouns referring to unique
denotata: John is John (OK, he has done that bad thing again, but you
know John, he's always like that; he has to be accepted as he is, because
he cannot be changed); with verbal tautologies of different kinds: If it's
raining it's raining, Whatever will be will be, and so on.
Perhaps the most plausible candidate for a common denominator of
the different tautological expressions commonly used in English is
something like this:
this cannot be changed
(or: this cannot be not like this)
This seems to fit expressions conveying distrust such as A snake is a
snake, as well as expressions conveying tolerance or indulgence such as
Kids are kids or Boys will be boys. It also seems to fit expressions
referring to the future, such as Whatever will be will be, those referring
to the past, such as Let bygones be bygones, and those referring to
activities, such as War is war. It also fits expressions like Enough is
enough, which indicate the speaker's determination not to budge, as well
as the expression That's that (whether it signals the same determination
not to budge or whether it refers to an irreversible event such as death).
442 Boys will be boys: even 'truisms' are culture-specific
Furthermore, it appears to fit well certain other types of sentences
which haven't thus far been mentioned in this chapter, such as
When you're out of form, you're out of form.
(a remark made by commentator Bill Lorry during a recent televised
cricket match, referring to the well-known cricket player, David Gower,
who had been having a bad run with the bat). Consider also the following
example from a recent newspaper article:
In the Brussels stadium we saw a display of lethal stupidity.
Meanwhile, the dead are dead and are past wondering why they
had to be killed. (The Weekend Australian 1-2.6.85)
In the context of this article, the tautological expression is neither
topic-closing nor dismissive. But the message 'this cannot be changed'
is clearly conveyed.
An added bonus of an analysis in terms of 'this cannot be changed'
is that it offers a possibility of spelling out something like the idea of
'acceptance' in a way which may seem to be applicable to all tautologi-
cal expressions, including those which convey distrust:
one should know: this cannot be changed
The frame 'one should know' combined with the component 'this cannot
be changed' permits us to express an idea akin to acceptance without
conveying connotations incompatible with any specific use to which the
tautological pattern is put in English.
Nonetheless I have not used the formula 'this can't be changed' in all
the explications proposed in this chapter, because I am not fully con-
vinced that it is the optimal one in each particular case, even if it is
the optimal 'compromise' between all the different constructions. For
example, if one says
I don't care what brand - coffee is coffee.
does one really mean 'this cannot be changed'? Doesn't one mean,
rather, something like 'it makes no difference - all Xs (all kinds of
coffee) are always the same' ?
On the face of it, the idea of 'immutability', of 'this cannot be
changed', seems also to fit rather well the three-part patterns, presuma-
bly derived from Gertrude Stein's famous sentence A rose is a rose is
a rose. For example, the composer Tristram Kerry, to a question about
works in an electronic mode, replied (on ABC-FM radio, 10.6.85): "A
Is there a semantic invariant? 443
composition is a composItIon is a composition". Presumably, in say-
ing so Kerry wished to emphasise the continuity between traditional
composition and electronic composition. The formula 'this cannot be
changed' seems to fit quite well, but does it really capture in the best
possible way the communicative import of the utterance? Wouldn't it
be more illuminating and more accurate if we spelt it out along the
following lines: 'in essence, all Xs are the same' or: 'no matter how Xs
differ from one another, there is something about them all that is the
same; and this is the most important thing about them'. Consider also the
following sentence:
It had been hard, for they were links of friendship rather than
authority which bound him, but 'chains are always chains, even
when made of flowers'. (Tolstoy 1983:258)
Is the idea here that 'this cannot be changed' ? Or is it, rather, that 'Xs
can be different from one another in some ways, but in essence they are
all the same' ?
Another plausible candidate for a semantic invariant could be formu-
lated along the following lines:
this cannot be denied
(one cannot say: no, it is not like this)
Possibly, the restriction on the use of modifiers in tautological construc-
tions (??Little boys will be little boys, ???A hasty, insincere kiss is a
hasty, insincere kiss) could be explained in these terms: if the predicate
repeats a bare noun, the identity of the subject and the predicate is
obvious, and so it must be obvious to everyone that the claim made by
the speaker must be valid. The use of a more complex noun phrase could
obscure the identity of the subject and the predicate, and so it would no
longer be 'obvious' that what the speaker is saying cannot be rejected.
The foregoing discussion of a possible semantic invariant of tautologi-
cal constructions can be summarised in the form of the following four
hypothetical components, which I propose as a starting point for further
investigation:
(a) I don't want to say more things about this ('dismissive
quality')
(b) it cannot be not like this ('immutability', 'it cannot be
changed')
444 Boys will be boys: even 'truisms' are culture-specific
(c) all things of this kind are the same ('identity')
(d) one cannot say: no ('undeniability')
For English nominal tautologies, a more specific set of four recurring
components can be tentatively proposed:
(a) everyone knows ...
(b) I know: someone can think ...
(c) I think: one should not think this
(d) one should know ...
I believe that the problem of a possible semantic invariant of tauto-
logical expressions of different kinds cannot be seriously studied without
a set of tentative semantic representations of individual expressions or
constructions. The formulae proposed in this chapter can be discussed,
criticised and revised. In the process, the problem of a hypothetical
semantic invariant will, I think, get a chance of being clarified, if not
actually solved.
6. The deceptive form of English tautological
constructions
I think one reason why English tautological constructions have been so
widely mistaken for universal devices fully interpretable on the basis
of Grice's maxim of quantity, is that their form is deceptively simple,
and deceptively similar to logical tautologies.
For example, Brown - Levinson (1978:225) draw a distinction be-
tween what they call 'conventionalised tautologies' and other tautologies
(presumably, non-conventionalised ones). 'Conventionalised tautologies'
are illustrated with the Tamil sentence:
Avaar-aam avaaru.
'He they say is he.' (i.e. 'big deal him')
By contrast, English tautologies such as:
War is war.
Boys will be boys.
The deceptive form of English tautological constructions 445
are treated as non-conventionalised. According to Brown and Levinson,
by uttering such tautologies, "5 [the speaker] encourages H [the
hearer] to look for an informative interpretation of the non-informative
utterance." (1978:225).
But aren't these English-language tautologies in fact just as 'con-
ventionalised' as the Tamil one? Is it not an ethnocentric illusion to
regard English tautologies as 'natural', readily interpretable, based on
universal maxims of conversational behaviour, rather than on conven-
tions of the English language? Isn't the difference between these two
utterances:
War is war. Uustification)
A war is a war. (obligation)
largely conventional? Aren't the differences between these expressions:
Boys will be boys.
Knaben bleiben immer Knaben. (lit. 'Boys remain always boys')
Oni ze mal' tiki. (lit. 'They PRT boys')
A father is a father.
Un pere est toujours un pere. (lit. 'A father is always a father';
cf. Bally 1952: 17)
due to language-specific grammatical conventions?
It is indeed remarkable that English nominal tautologies tend to use
the very simple formal pattern: N is N (with only minor variations),
similar to logical formulae such as p is p. Other languages often use
more complex patterns, which include particles, modal verbs, adverbs
of time such as immer or toujours ('always') and so on. But is the
English formal pattern really more 'natural', less conventionalised, than
the complex patterns used by other languages? And isn't it the case
that, despite appearances, English, too, has a number of different
formal patterns, associated with different interpretations (N is N, A N is
a N, Ns are Ns, Ns will be Ns, and so on)?
If the analyses contained in this chapter do something to counteract
the spread of such illusions, its purpose will have been achieved.
446 Boys will be boys: even 'truisms' are culture-specific
7. The culture-specific content of tautological patterns
Tautologies are like gestures or interjections: they seem 'natural', uni-
versal, and self-explanatory; and they are all that, to some extent; but
they are also conventional, language-specific, and culturally revealing.
They manifest both 'natural logic' and 'cultural logic'; and the more
'natural' they appear the more culturally revealing they are.
Consider, for example, the gesture of 'nose-tap', widespread in
Europe and probably elsewhere as well (see Morris et al. 1979). The
nose is naturally associated with smell, and with sniffing, and sniffing
can be seen as naturally associated with alertness and with special
knowledge. It seems entirely 'natural', therefore, that the nose-tap
gesture is indeed associated with meanings of this kind. Nonetheless,
the nose-tap does not mean the same in different parts of Europe. In
English-speaking and Italian-speaking areas, it is used mainly to indicate
a shared secret (with two peaks occurring in Scotland and Sardinia). But,
as Morris et al. (1979:220) note, "almost as popular... is the meaning of
friendly warning ['Be alert! ']. This was almost non-existent in the north,
but was strongly favoured on the Italian mainland, with the peak at
Rome and Naples." In addition to this friendly meaning, an unfriendly
one - 'You are nosey!' - has also emerged in parts of Europe. "As an
accusation of interference, the gesture was almost entirely confined to
the English-speaking region, with only two isolated cases elsewhere."
(1979:224).
This is a perfect illustration of the way how culture-specific concerns
make use of 'natural' and universally-available means. In Italian culture,
group solidarity and interpersonal involvement play a greater role than
desire for personal autonomy, non-imposition, and 'non-interference',
whereas in the Anglo-Saxon culture the latter values are known to play
a dominant role. The sign in question (the nose-tap) is seen everywhere
as 'natural' and self-explanatory; but in fact, its interpretation is
shaped partly by the shared cultural traditions, and cultural preoccupa-
tions, of a particular community of speakers.
Exactly the same applies to tautologies. They are like convenient all-
purpose vessels, into which 'basic truths' can be poured - but 'basic
truths' are different in different cultures.
Consider, for example, the following description of an American
Jewish wedding, in a novel by Bernard Malamud:
The culture-specific content of tautological patterns 447
The guests, including the notables, rise, lift their feet, and dance . ... The
women serve a feast of chicken with sesame and tomatoes, roasted yams,
and palm wine. ... Those who feel like crying, cry. A wedding is a wedding.
(Malamud 1971:217)
The form of the tautology in the last line is impeccably English, but its
use in this passage is slightly odd, because it doesn't fit into any of the
semantic categories which have evolved in the English language (such
as absolute generalisation, indifference, obligation, or undeniable value).
Apparently, what Malamud is trying to convey is this: 'human life is
complex, impenetrable, uncontrollable; its joys must be interwoven with
sorrows; it is natural for people to cry, as well as to dance and to eat, at
a wedding; such is our experience of life.' But this is Jewish wisdom,
embodied in particular in the Yiddish language and lore (cf. Matisoff
1979), not Anglo-Saxon wisdom. As a result, Malamud's tautology
sounds like a calque, and a cultural loan, from Yiddish.
The same applies to another tautological expression, used by the same
writer in the same context. At the same wedding the rabbi addresses the
newlyweds in the following words:
Now you are man and wife ... I feel like crying, but why should I cry if the
Lord says, "Rejoice!" Willie and Irene, listen to me. Oh, what a hard thing
is marriage in the best of circumstances. ... All I am saying is the world is
imperfect. ... Willie and Irene, to enjoy the pleasures of the body you
don't need a college education; but to live together in love is not so
easy. Besides love that which preserves marriage is that which preserves
life; this is mutual trust, insight into each other, generosity and also
character, so that you will do what is not easy to do when you must do it.
What else can I tell you, my children? Either you understand or you don't.
(Malamud 1971:215-216)
Again, at first sight the verbal tautology in the last line seems impecca-
bly English; and yet this time, too, there is something slightly odd about
it. In English, v e ~ l tautologies of the 'either-or' type refer, normally,
to actions, not to understanding. One can of course say in English:
"What do you mean, you're not sure? Either you know or you don't"
(either you understand or you don't); but in this case the speaker would
be urging the addressee to make a choice. Except for such promptings,
however, an 'either-or' English tautology would normally refer to an
action. (Either he comes or he doesn't). The point of such tautologies
was spelt out earlier: 'one shouldn't waste one's time and energy in
useless worry; if one can't change something there is no point in fretting
448 Boys will be boys: even 'truisms' are culture-specific
over it; it is better to turn one's mental energy to something more
productive'. This is perfectly in keeping with the pragmatism of
Anglo-Saxon culture.
But the point of Malamud's tautology is different. The rabbi's
experience of life is sad ("The world is imperfect"). In uttering his
tautology he is not trying to dismiss his sadness and melancholy. He is
resigned to the futility of his own talk, and he recognises its pointless-
ness, but he is not trying to spare himself from 'uselessly feeling some-
thing bad'. (After all, Jews are a people who have for centuries, indeed
millennia, remembered, and 'uselessly' lamented in their midnight
service, the destruction of their temple two thousand years ago. 'Either
you understand or you don't ... '.)
One might argue, of course, that Malamud's tautologies are 'culturally
odd', not 'linguistically unacceptable'. I wouldn't want to say that, but
this is not the point: I have presented in this chapter many examples
of tautologies whose meanings couldn't possibly be 'worked out' by
speakers of other languages, and which are, therefore, indubitably
language-specific, as well as culture-specific. But no matter how one
categorises the slight 'deviance' of Malamud's tautologies, they illus-
trate very well the way how certain formal patterns, which are felt (in a
particular culture) to embody unquestionable universal wisdom, can in
fact specialise in expressing wisdom which is highly culture-specific.
What is 'self-evident' in one culture may be questionable or unaccept-
able in another. What is seen as a 'fundamental truth' in one culture may
be seen as not worth saying in another. What is seen as an 'unquestiona-
bly wise attitude' in one culture might be seen as foolishness in another.
Perhaps this is, ultimately, the sense of all (most?) tautologies, in all
(most?) languages: that they - like proverbs - are supposed to embody
'wisdom'; and, in particular, that they recommend certain attitudes to
life as 'unquestionably wise' and as tested by generations of people.
8. Conclusion
In describing a specific language (for example, English) as a system,
somewhere or other a line will have to be drawn between what is in
and what is out, between grammar and 'human nature', between meaning
and 'implicature'. Does it matter very much where exactly this line is
Conclusion 449
drawn? Does it matter, for example, if constructions such as Boys will be
boys are counted as being 'in' or 'out', i.e. as part of English grammar,
whose meaning is language-specific and has to be included in the
linguistic description of English as an individual language, or as free
collocations whose use is 'calculable' from general principles of
human conversation?
In my view, it is a matter of fundamental importance. Obviously, this
importance doesn't derive from the frequency of use, or the indispensa-
bility, of 'tautological sentences' as such (although their cultural
importance may be greater than one would initially suspect). It derives
from the fact that one's decision on a 'small' point like this has far-
reaching consequences with respect to one's whole idea of what linguis-
tics is all about, what it is supposed to do, and what it can do. The basic
question is this: should grammar be autonomous of pragmatics, or
should 'pragmatic meanings' (i.e. some matters to do with the speakers'
assumptions, intentions, thoughts and feelings) be accounted for in the
same over-all descriptive framework which is used for 'objective' or
truth-functional aspects of language use?
Many contemporary linguists seem to feel that they are caught in an
insoluble dilemma. On the one hand, it has been demonstrated again
and again that 'pragmatic' (or 'subjective') and 'objective' aspects of
language use are interrelated and that grammar interacts with illocution-
ary force and other 'pragmatic' matters. Furthermore, consistent attempts
to separate grammar from 'pragmatic meanings' lead to the paradoxical
conclusion that a good deal of conversational English is 'ungrammati-
cal'. For example, perfectly ordinary and perfectly acceptable sentences
such as Would you please be quiet?, or Why don't you be quiet! have
been classified as 'ungrammatical' - simply because 'autonomous'
grammar cannot possibly account for their use. (See for example Bach
- Harnish 1979.) This outcome would seem to represent a reductio
ad absurdum of the autonomous (non-illocutionary) grammar stand.
On the other hand, the most influential attempt to develop a frame-
work for an integrated description of grammar and 'pragmatic' phenom-
ena such as illocutionary force, the approach associated with the
generative semantics school, has ended in self-acknowledged defeat.
This has resulted in widespread disillusionment not only with generative
semantics but also with any attempts to integrate a grammatical and a
pragmatic description of language.
What is to be done, then? To many linguists, the only salvation
seems to be offered by a Gricean approach to language. This state of
450 Boys will be boys: even 'truisms' are culture-specific
mind can perhaps be expressed as follows: 'Somebody has to account for
language use, but we linguists have now come to realise that we cannot
do it. Fortunately, we don't have to feel guilty about it any longer. We
now see that it is simply not our responsibility. Another science will do
it: a science of human behaviour in general and human conversational
behaviour in in particular. Linguists can concentrate on studying
language structure. And in fact we can now say that it would be a
mistake to confuse language structure with language use. Grammar is
one thing, and illocutionary force another; meaning is one thing and
'implicature' another. We neither can nor should try to account for
language use. For example, we don't have to try to predict the communi-
cative import of sentences such as Boys are boys or Business is business.
Certainly, such matters are not devoid of interest, but they will be taken
care of by the science of conversational behaviour.'
In Sperber's - Wilson's words (1981:296, on a slightly different but
related matter), "Grice's proposal would relieve semantic theory of the
problems of defining" pragmatic meanings. After the self-proclaimed
collapse of generative semantics, which tried to develop a framework for
a semantic analysis of illocutionary meanings, contemporary semantics
seems to have lost all confidence in itself. A philosopher who comes and
apparently 'relieves the linguist' from the obligation of analysing many
kinds of troublesome meanings, and in particular, illocutionary mean-
ings, tends to be treated like a saviour, and to be accorded total faith.
In my view, this faith in the omnipotence of Gricean maxims is bound
to be disappointed. Neither tautological constructions nor numerous
other matters which have been left to the postulated new science based
on Gricean maxims, will be automatically taken care of. Ultimately,
numerous 'tautological sentences' used in conversational English will
simply have to join the ever-growing pile of rejects - i.e. of sentences
which are perfectly acceptable but which are 'ungrammatical', because
neither Gricean maxims nor an autonomous grammar can account for the
limits of their use. (For example, why can one say Boys will be boys but
not *War will be war or *Boys might be boys? Why can one say Kids
are kids but not ?Bottles are bottles or *Clouds are clouds? Why can
one say Enough is enough but not *Much is much? Why can one say
A bet is a bet but not *Bets are bets?
Of course nobody denies these days that pragmatic information is
reflected in some aspects of grammar, and that, consequently, some as-
pects of grammar cannot be fully accounted for without some reference
to the speaker's and hearer's assumptions, feelings, and so on. But so
Conclusion 451
long as it seems possible to regard such reflections as minor and isolated,
they can be ignored by the grammarian and the ideal of autonomous
grammar can be upheld (even if it has to be done at the cost of discarding
some acceptable sentences as 'ungrammatical '). I would argue, however,
that pragmatics pervades grammar, that the two intertwine in countless
areas, and that the bulk of conversational English can be explained only
on the basis of this interaction. (Cf. Bally 1952.)
Given that 'conversational meanings' (Gazdar's term, 1979:55) cannot
be explained on the basis of universal principles of conversation, can
they perhaps be accounted for on the basis of some language-specific
science of language use, separate from grammar (as suggested by
Morgan 1978)?
If we followed this line of reasoning we could say that the use of
tautological constructions is determined partly by conventions which are
specific to English, but that these conventions are not 'grammatical' in
nature, and can be kept apart from a linguistic description of English
structure.
To my mind, this program is no more realistic than Grice's (or rather,
than the program advanced by the linguistic followers of Grice, who
often tend to be more Gricean than Grice himself). Attitudinal meanings
enter the core of grammar. If we establish any correlation between the
imperative (Do it!) and the meaning 'I want you to do it', or between the
interrogative structure and the meanings 'I want to know it' or 'I want
you to say it', or between tag questions and meanings such as 'I think X,
I assume you think the same, I want you to say it', or between exclama-
tions and meanings such as 'I feel something', we are already 'mixing'
grammar and pragmatics. But can any empirically adequate grammar
ignore such correlations?
Once it is recognised that a semantic metalanguage capable of repre-
senting both 'objective' and 'subjective' aspects of meaning constitutes a
sine qua non of linguistic description then subtle pragmatic meanings
like those encoded in tautological constructions cease to present a seri-
ous problem. The linguist can recognise that such constructions are at
least partly language-specific, and can seek to account for their use with-
out appealing to any deus ex machina in the form of extra-grammatical
linguistic conventions: it becomes clear that the relevant meanings
can be modelled in the same semantic metalanguage as all other kinds
of meaning.
452 Boys will be boys: even 'truisms' are culture-specific
Furthermore, once the language-specific meanings of various
'tautological' constructions are spelt out in a rigorous (and yet self-
explanatory) semantic metalanguage, it becomes possible to study
universal tendencies and regularities in the use of such constructions,
which are no less real, and no less accessible to precise semantic
analysis, than the language-specific uses to which such constructions
can be put.
Theories of conversation (such as Grice's) can then proceed without
any unrealistic burdens being placed on their shoulders. 'Ungrammati-
cal' sentences such as Why don't you be quiet or Would you please be
quiet can be rehabilitated as fully grammatical encodings of language-
specific pragmatic meanings. An integrated theory of linguistic descrip-
tion can, once more, be set out as a goal of linguistics - and as a
responsibility of which no other sciences can relieve it.
Chapter 11
Conclusion: selDantics as a key to
cross-cultural pragmatics
Language is, it is said, a tool of communication. As a statement, this is of
course true, but as a definition, it is narrow, one-sided, and inadequate.
Language is indeed a tool of communication, but - crucially - it is
also a tool of human interaction. By means of language, we express
our personality, our thoughts, intentions, desires, and feelings; and by
means of language we relate to other people.
This book develops a framework for studying language as a tool
of human interaction. 1 have tried to show that we interact with other
people by expressing certain meanings - and 1 have developed a
method for revealing and stating these meanings in a clear, simple,
and rigorous way.
For example, modes of interaction associated with English words
such as reprimand, praise, thank, apologise, threaten, promise, warn,
and complain have been analysed by means of simple and intuitively
understandable semantic ~ o p o n n t s such as the following ones:
you did something bad
someone did something good
you did something good for me
1 did something bad to you
1 will do something bad to you
1 will do something good for you
something bad may happen to you
something bad happened to me
(reprimand)
(praise)
(thanks)
(apology)
(threat)
(promise)
(warning)
(complain)
Components of this kind are formulated in a kind of simple, reduced
English, and they can be immediately understood (via English). But
unlike the great majority of 'normal' English sentences they could be
readily translated into other languages, because they rely on very
simple grammatical patterns, and on universal, or near-universal,
human concepts such as 'you' and 'I', 'good' and 'bad', or 'do' and
'happen', which have their counterparts in all, or nearly all, languages
of the world.
454 Conclusion: semantics as a key to cross-cultural pragmatics
Components of this kind, formulated in the language of universal
'semantic primitives', allow us to study human interaction in a system-
atic and rigorous way, without relying on any technical jargon or on
mathematical and logical models, which, when employed in the study
of language and language use, often give the impression of scientific
rigour but in fact obscure rather than clarify the phenomena under
consideration.
Human interaction carried out by means of language is shaped and
coloured to a very considerable extent by simple meanings such as
'you did something bad/good' or 'I did something bad/good', 'I want
something' or 'I don't want this', and so on; but meanings of this kind
are obscured, not clarified, when human interaction is discussed in
terms of 'negative' or 'positive face' (cf. Brown - Levinson 1978), or
'cooperativeness' (cf. Grice 1975) or 'indirectness' (cf. Searle 1975).
Above all, complex terms of the kind usually employed in studies of
linguistic interaction are derived from English (academic English), and
don't have exact counterparts in other languages of the world. As a
result, the description of human interaction carried out by means of
such labels is usually remarkably ethnocentric (anglocentric). This is
in keeping with the general anglocentric bias of mainstream modern
pragmatics, which developed largely under the influence of British and
American philosophers of language, who drew their examples almost
exclusively from English and who postulated all sorts of 'universal
maxims' or 'universal principles' of human interaction on the basis of
Anglo-Saxon or Anglo-American cultural norms and expectations.
The present book is based on a wide range of data, drawn from a
wide range of languages, including English, Italian, Russian, Polish,
Yiddish, Hebrew, Japanese, Chinese, Korean, Walmatjari (an Australian
Aboriginal language), and many others. It also draws on different
varieties of English, including Australian English and Black American
English. It rejects 'universal maxims' of human behaviour based on
mainstream British or American English, and it refuses to treat
descriptive categories based on academic English as a 'culture-free'
and language-independent analytical framework.
Instead, it relies on a 'natural semantic metalanguage', based on
universal or near-universal semantic primitives; and it employs this
metalanguage to explore and to analyse a wide variety of cultural norms
and cultural traditions from a universal, language-independent perspec-
tive. It recognises the enormous variety of 'communicative styles'
linked with different human languages and refuses to treat this variety
Conclusion: semantics as a key to cross-cultural pragmatics 455
in terms of deviations from one 'basic', 'natural' model (that associated
with mainstream American, or British, English). At the same time, it
recognises the underlying unity of human interaction, determined by the
universality of concepts such as 'you' and 'I', 'do' and 'happen', 'good'
and 'bad', or 'want' and 'know'; and by the universality of semantic
components such as 'I want something', 'I know something', 'you know
something', 'I don't want this', 'I/you/someone did something bad', 'I
don't want this', 'I want you to say something', 'I want you to do
something', and so on.
Different communicative styles, and different norms of social interac-
tion, are described here in terms which are not only universal (language-
and culture-independent), but also simple and easy to understand, to the
point of being accessible to children, and to second-language learners
(for example, immigrants) with a very limited knowledge of English.
It can be hoped, therefore, that the kind of simple metalanguage used
here to investigate different modes of human interaction and different
interactive strategies associated with different languages can be used
not only as a tool for investigating linguistic interaction in different
cultural settings, but also as a basis for teaching it; and in particular, as
a basis for teaching successful cross-cultural communication.
Notes
1. The arrow (= indicates an implication, which forms a part of the full
meaning of the expression.
2. The key place of forms of address in cross-cultural pragmatics is highlighted
by Friederike Braun's (1988) valuable study, Terms of address: problems
of patterns and usage in various languages and cultures, which came to my
attention after the completion of the present work.
3. Baxtin's study "Problema recevyx zanrov" ['The problem of speech genres']
was published in 1979, but actually written in 1952, so it seems to me that
he can be rightly seen as a precursor of the modem Western literature on
speech acts and speech genres which flowered in the 1960s and 1970s. In
fact, Baxtin's first work on related topics was published under the assumed
name of Volosinov as early as 1929. But the 1929 book, interesting and
original as it is, does not focus on speech genres, as does the 1952 study.
For an English translation of this pioneering work, see Baxtin (1986). It is
devoted specifically to speech genres, and it outlines a program which I
find particularly congenial.
4. Since most contemporary linguistic writings on speech acts seem to assume
that this idea is a modern invention, I think it is in order to offer here a
lengthy quotation which will help put the matter in a more realistic historical
perspective:
Starting from the domain of dialectic and probably utilising already
existing divisions of kinds of speech the Stoics developed rather
elaborate classifications of complete lekta. From the several passages
in which such lists are given we learn that apart from axiomata and
the two kinds of questions the following varieties of complete lekta
were distinguished.
(a) That which is more than an axioma or is like an axioma, for example:
'How beautiful the Parthenon (is)', 'How like to Priam's sons the
cowherd (is)'. It differs from an axioma in so far as it includes some
emotional reaction, usually expressed by the particle hos ('How'). It
is also called thaumastikon, the lekton of admiration or astonishment;
as such it may be contrasted with the psektikon, the lekton of censure.
(b) Another sort of question, put to oneself in order to express doubt or
despair (epaporetikon): 'Are sorrow and life something akin?'.
(c) A command (prostaktikon), for instance: 'Come thou hither, 0 lady
dear'.
458 Notes
(d) An oath (horkikon, omotikon, epomotikon), for instance: 'By this
sceptre', 'Let earth be my witness in this'.
(e) A prayer (aratikon, euktikon) , for example: 'Zeus, my father, who
rulest from Ida, majestic and mighty, victory grant unto Ajax and
crown him with glory and honour'. Sometimes a distinction is drawn
between prayer (euktikon) and deprecation or curse (aratikon); an
instance of the latter is 'Even as this wine is spilt, so may their brains
be spilt earthwards'.
(f) An address or greeting (prosagoreutikon), for example: 'Most
honoured son of Atreus, lord of the warriors, Agamemnon'.
(g) A supposition or assumption (hypothetikon), for instance: 'Let it be
supposed that the earth is the centre of the sphere of the sun'.
(h) An exhibition of a particular instance (ekthetikon): 'Let this be a
straight line'. According to DL VII, 196, Chrysippus had written a
treatise about ektheseis.
(i) An explanation or elucidation (diasaphetikon).
(Nuchelmans 1973:63)
5. This earlier version of the performative analysis is also described by
Nuchelmans:
If one utters the sentence '0, that the king might come' (utinam rex
veniret), the composite thought produced in the hearer, the intellectus
of that sentence, is the same as that produced by the sentence 'I wish that
the king would come' (volo regem venire or opto, ut rex veniat). If one
utters the sentence 'Help me, Peter' (adesto, Petre), either as a command
or as a prayer, the intellectus is the same as that of the sentence 'I order
you to help me' (praecipio, ut adsis mihi) or of the sentence 'I pray you
to help me' (deprecor, ut adsis mihi). The utterance 'Peter!' (0 Petre)
produces the same mental counterpart as the sentence 'I call you, Peter'.
And the question 'Did Socrates come?' produces the same thought as the
sentence 'I ask you if Socrates came'. Similarly, the one-word utterances
papae and heu produce the same thought as ego admiror ('I admire (it)')
and ego doleo ('I feel grief (at it) '), although in the first two cases the
thought is non-composite and in the second two cases the thought is
composite.
This means that in spite of the differences in force of the vocal
utterances, namely between the non-declarative utterance and the state-
ment-making utterance which in each case corresponds to it, there is no
difference between the mental counterparts of the non-declarative
utterance and the corresponding statement-making utterance. The
composite thought produced in the hearer's mind by each of the several
non-declarative utterances is the mental counterpart of the propositio
that may take its place. (Nuchelmans 1973:148)
Notes 459
6. Examples drawn from popular fiction and plays are indicated as follows:
AC: Anna Christie (O'Neill 1960)
GH: The great Gillie Hopkins (Paterson 1978)
GP: The glittering prizes (Raphael 1977)
NI: The night of the iguana (Williams 1961)
WE: Welded (O'Neill 1958)
7. Felix Ameka (p.c.) has suggested that 'vocal gestures' such as YukI should
be distinguished from 'speech acts' such as I feel disgusted along the
following lines:
I feel disgusted.
I say: I feel disgusted
I want someone to know it
I say this because of that
YukI
I feel disgusted
I want someone to know it
I do this: [vocal gesture] because of that
I find this an interesting proposal, which requires further investigation.
8. For earlier attempts at semantic analysis of some interjections based on a
controlled semantic metalanguage, see Ameka (1987; to appear) and Hill
(1985); see also Wierzbicka (1986a).
9. The expression 'I now know' refers to something that has just come to the
speaker's attention. It might seem more natural to use the expression 'I
perceive' instead. But perceive, unlike know, cannot be presumed to be an
indefinable or a lexical universal, and consequently know must be preferred
to it, as an element of the semantic metalanguage.
10. Some readers have objected to the suggestion that wow necessarily implies
'something good'. The counterexamples they have offered, however, could
all be interpreted as sarcastic or ironic, that it, as relying for their effect
on an underlying presumption of 'good'. For example, a parent may say
Wow! in response to either I got three As today! or I got three Ds today!,
but in the latter case such a response as Wow! would normally be interpreted
as sarcastic.
11. The range of use of yuk is not uniform throughout the English-speaking
world; it is particularly widely used in Australia.
12. Consider also the following dialogue from a play by a leading Australian
playwright:
Don: (to Mal) Why in the hell do you get her pregnant every eighteen
months?
Mal: She won't take the pill. ...
Don: (to Mal) Use Silvertex.
460 Notes
Mal: (with distaste) YukI
(Williamson 1973:210)
13. As pointed out to me by Professor Werner Winter (p.c.), Latin had a special
dictum corresponding to the English 'tautologies of human nature': sunt
pueri pueri; pueri puerilia tractant, literally '(male) children are (male)
children; children do childish things'. As Professor Winter points out, this
dictum makes the relationship between members of the class and their
prototypical behaviour beautifully explicit. At the same time, the word
puerilia 'childish (things)', clearly implies an indulgent attitude, incompat-
ible with serious condemnation.
Bibliography
Abrahams, Roger
1970 "Rapping and capping: Black talk as art", in: Szwed (ed.), 132-
142.
1974 "Black talking on the streets", in: Bauman - Sherzer (eds.),
240-262.
1976 Talking black (Rowley, Mass.: Newbury House).
Ameka, Felix
1986 The use and meaning of selected particles in Ewe (M.A. thesis,
Australian National University).
1987 "A comparative analysis of conversational routines in two
languages: English and Ewe", Journal of Pragmatics 11 :299-326.
to appear "Agoo can be English, can it? On the semantic representation
of linguistic routines", in: David Wilkins (ed.), Conceptual
primitives and semantic analysis: A cross-linguistic perspective
(Berlin: Mouton de Gruyter).
AND
1988
"Comparative adjectives and adverbials of degree: an introduction
to radically radical pragmatics", Linguistics and Philosophy
7:347-377.
Austin, John L.
1962 How to do things with words (Oxford: Oxford University Press).
Bach, Kent - Robert Harnish
1982 Linguistic communication and speech acts (Cambridge, Mass.:
MIT Press).
Baker, Sidney J.
1959 The drum: Australian character and slang (Sydney: Currawong).
1966 The Australian language (Sydney: Currawong).
Bakhtin, see Baxtin
Bally, Charles
1952 Le langage et la vie (3rd expanded edition) (Geneve: Droz).
The Australian national dictionary: a dictionary of Australianisms
on historical principles (W.S. Ramson, ed.) (Melbourne: Oxford
University Press).
Arensburg, Conrad M. - Arthur H. Niehoff
1975 "American cultural values", in: Spradley - Rynkiewich (eds.),
363-377.
Ash, see Garton Ash
Atlas, Jay
1984
462 Bibliography
Bankowski, Andrzej
1971 "Jeszcze i jui" [Jeszcze andjui], J(lzyk Polski 51:21-30.
1975a "0 kwantyfikacji i partykule az" [On linguistic quanti-
fication and the particle az], Prace Filologiczne 25:183-189.
1975b "Jeszcze przy komparatywach" [Jeszcze with comparatives],
Poradnik 3:148-150.
1976 "Opozycja semantyczna partyku} dopiero i jui" [The semantic
opposition of the particles dopiero and jui], Prace Filologiczne
26: 13-38.
1977 "Opozycja semantyczna partykul jeszcze i jui" [The semantic
opposition of the particles jeszcze and jui], in: Siatkowski (ed.),
7-46.
Barnlund, Dean
1975a "Communicative styles in two cultures: Japan and the United
States", in: Kendon - Harris - Key (eds.), 427-456.
1975b Public and private self in Japan and the United States: communi-
cative styles of two cultures (Tokyo: Simul).
Bartlett, John
1980 Familiar quotations (15th edition) (London: Macmillan).
Barzini, Luigi
1964 The Italians (London: Hamish Hamilton).
Bates, Elizabeth
1976 Language and context: the acquisition of pragmatics (New York:
Academic Press).
Bauman, Richard - Joel Sherzer (eds.).
1974 Explorations in the ethnography of speaking (Cambridge:
Cambridge University Press).
Baxtin, Mixail
1979 "Problema recevyx zanrov" [The problem of speech genres] [Writ-
[1952-3] ten in 1952-3, published in his Estetika slovesnogo tvorcestva
1979, 237-280 (C.G. Boxarov, ed.)] (Moskva: Iskusstvo).
1986 Speech genres and other late essays [Mikhail Bakhtin] (Vern W.
McGee, trans.) (Austin: University of Texas Press).
Benedict, Ruth
1947 The chrysanthemum and the sword: patterns of Japanese culture
(London: Seeker & Warburg).
Blount, Ben (ed.)
1974 Language, culture and society (Cambridge, Mass.: Winthrop).
Blum-Kulka, Shoshana
1982 "Learning to say what you mean in a second language: a study of
the speech act performance of learners of Hebrew as a second
language", Applied Linguistics 3:29-59.
Bibliography 463
1989
1988
1972
1986
1975
1981a
1981b
Blum-Kulka, Shoshana - Elite Olshtain
1984 "Requests and apologies: a cross-cultural study of speech act
realization patterns", Applied Linguistics 5: 196-213.
Blum-Kulka, Shoshana - Brenda Danet - Rimona Gherson
1985 "The language of requesting in Israeli society", in: Forgas (ed.),
113-139.
Blyton, Enid
1980 The mystery of the Banshee Towers (London: Granada).
Bogardus, Emory S.
1933 "A social distance scale", Sociology and Social Research 17:
265-271.
Boguslavskij, Igor' M.
1985 Issledovanija po sintaksiceskoj semantike. Sfery dejstvija
logiceskich slov [Investigations into syntactic semantics: the
spheres of action of logical words] (A.P. Ersov, ed.) (Moskva:
Nauka).
Bogustawski, Andrzej
1966 Semantyczne pojfcie liczebnika [The semantic concept of the
numeral] (Wroclaw: Ossolineum).
"0 wyrazeniach [On expressions denoting
change], in: Wierzbicka (ed.), 147-149.
"On 'the world' ", Linguistica Silesiana 1:63-69.
"Semantic and pragmatic aspects of reference related problems",
in: Danes - Viehweger (eds.), 1-111.
"Wissen, Wahrheit, Glauben: zur semantischen Beschaffenheit
des kognitiven Vokabulars", in: Bungarten (ed.), 54-84.
"Also from all so: on a set of particles in service of efficient
communication", Journal of Pragmatics 10:615-633.
"Glosa do akt6w mowy" [A gloss to the book on speech
acts], Pamiftnik Literacki 79:103-124.
"Knowledge is the lack of lack of knowledge, but what is that
lack lack of?", Quaderni di Semantica 10.1: 15-31 (Round table
on semantic primitives, 1).
Bogus}awski, Andrzej - Bozena Bojar (eds.)
1986 Od kodu do kodu [From code to code] (Warszawa: Panstwowe
Wydawnictwo Naukowe) (Festschrift for Olgierd Wojtasiewicz).
Bolinger, Dwight
1972 Degree words (The Hague: Mouton).
1977 Meaning and form (London: Longman).
1982 "Intonation and its parts", Language 58:505-533.
Bowles, Colin
1986 G'day! Teach yourself Australian in 20 easy lessons (Sydney:
Angus & Robertson).
464 Bibliography
Braun, Friederike
1988 Terms of address: problems of patterns and usage in various
languages and cultures (Berlin: Mouton de Gruyter).
Brown, Penelope - Stephen Levinson
1978 "Universals in language usage: politeness phenomena", in: Goody
(ed.), 56-310.
1987 Politeness: some universals in language usage (Cambridge:
Cambridge University Press).
Brown, Roger - Albert Gilman
1972 "The pronouns of power and solidarity", in: Giglioli (ed.), 252-
282.
Brzechwa, Jan
1983 Brzechwa dzieciom [Brzechwa for children] (Warszawa: Nasza

BUhler, Karl
1933 Sprachtheorie (Jena) [Reproduced 1965, Stuttgart: Fischer].
Bungarten, Theo (ed.)
1981 Wissenschaftsprache: Beitriige zur Methodologie, theoretischen
Fundierung und Deskription (MUnchen: Fink).
Butwin, Frances
1958 Translator's introduction to Sholom Aleichem [Rabinowitz], The
old country (London: Andre Deutsch).
Buzo, Alexander
1974 "The front room boys", in: Four Australian plays (Ringwood,
Vic.: Penguin).
1979 Norm and Ahmed, Rooted, The Roy Murphy Show (Sydney:
Currency Press).
Caton, Charles (ed.)
1963 Philosophy and ordinary language (Urbana, Ill.: University of
Illinois Press).
Cattell, Ray
1973 "Negative transportation and tag questions", Language 49:612-
639.
Chisholm, William S.
1984 Interrogativity: a colloquium on the grammar, typology and
pragmatics of questions in seven diverse languages (Amsterdam:
John Benjamins).
Chomsky, Avram Noam
1957 Syntactic structures (The Hague: Mouton).
1975 Reflections on language (New York: Pantheon).
Clancy, Patricia
1986 "The acquisition of communicative style in Japanese", in: Schief-
felin - Dchs (eds.), 213-250.
Bibliography 465
Clark, Herbert - Dale Schunk
1980 "Polite responses to polite requests", Cognition 8:111-143.
Clyne, Michael (ed.)
1976 Australia talks: essays on the sociology of Australian immigrant
and Aboriginal languages (Canberra: Australian National Univer-
sity) (Pacific Linguistics, D23).
Cole, Peter (ed.)
1978 Syntax and semantics 9: Pragmatics (New York: Academic Press).
1981 Radical pragmatics (New York: Academic Press).
Cole, Peter - Jerry Morgan (eds.)
1975 Syntax and semantics 3: Speech acts (New York: Academic
Press).
Coleman, Peter (ed.)
1962 Australian civilisation: a symposium (Melbourne: Cheshire).
Comrie, Bernard
1984a "Russian", in: Chisholm (ed.), 7-46.
1984b [Plenary session discussion], in: Chisholm (ed.), 255-287.
Condon, John C. - Mitsuko Saito (eds.)
1974 Intercultural encounters with Japan: communication - contact
and conflict (Tokyo: Simul).
Conway, Ronald
1971 The great Australian stupor: an interpretation of the Australian
way of life (Melbourne: Sun Books).
Cooke, Joseph R.
1968 Pronominal reference in Thai, Burmese, and Vietnamese
(Berkeley: University of California Press).
Coulmas, Florian
1981 "Poison to your soul: thanks and apologies contrastively viewed",
in: Coulmas (ed.), 69-91.
Coulmas, Florian (ed.)
1981 Conversational routine: explorations in standardized communica-
tion situations and prepatterned speech (The Hague: Mouton).
1986 Direct and indirect speech (Berlin: Mouton de Gruyter).
Cowie, Anthony P. - Ronald Mackin
1976-83 Oxford dictionary of current idiomatic English 1-2. (Oxford:
Oxford University Press).
Crawford, Raymond Maxwell
1970 Australia (London: Hutchinson).
Dal', Vladimir
1955 Tolkovyj slovar' zivogo velikorusskogo jazyka [Explanatory
[1882] dictionary of the contemporary Great Russian language] 1-4.
1974
Doi, Takeo
1973
466 Bibliography
(Moskva: Gosudarstvennoe izdatel' stvo inostrannyx nacion-
al 'nyx slovarej).
Danes, Frantisek - Dieter Viehweger (eds.)
1981 Pragmatische Komponenten der Satzbedeutung (Berlin: Akademie
der Wissenschaften der DDR) (Linguistische Studien Reihe A,
91/1).
Darwin, Charles
1955 The expression of the emotions in man and animals (New York:
[1874] Philosophical Library).
Davies, Norman
1981 God's playground: a history of Poland in two volumes 1-2.
(Oxford: Clarendon Press).
1984 Heart of Europe: a short history of Poland (Oxford: Clarendon
Press).
Davison, Alice
1975 "Indirect speech acts and what to do with them", in: Cole -
Morgan (eds.), 143-185.
Deakin, Greg
1981 Indirect speech acts and intonation (M.A. thesis, Australian
National University).
de Tocqueville, Alexis
1953 Democracy in America (Henry Reeve, trans.) (New York: Alfred
[1835-40] A. Knopf).
DeVos, George
1985 "Dimensions of the self in Japanese culture", in: Marsella -
DeVos - Hsu (eds.), 141-184.
Dixon, Robert M.W.
1977 "Where have all the adjectives gone?", Studies in Language 1: 19-
80 [Reprinted in his Where have all the adjectives gone? and
other essays in syntax and semantics, 1982, 1-62] (The Hague:
Mouton).
Doherty, Monika
1973 " 'Noch' and 'schon' and their presuppositions", in: Kiefer -
Ruwet (eds.), 154-177.
The anatomy of dependence (John Bester, trans.) (Tokyo: Kodan-
sha).
"Some psychological themes in Japanese human relationships",
in: Condon - Saito (eds.), 17-26.
Doroszewski, Witold
1938 ~ z y polski w Stanach Zjednoczonych [The Polish language in
the United States] (Warszawa: Naklad Towarzystwa Naukowego
Warszawskiego).
Bibliography 467
Dostoevskij, Fedor
1976 Brat'ja Karamazovy [The brothers Karamazov] (Polnoe sobranie
socinenij [Complete collected works], 14) (Leningrad: Nauka).
Dostoevsky, Fedor
1974 The brothers Karamazov (Constance Garnett, trans.) (London:
Heinemann).
Dovlatov, Sergej
1977 "Krug rasskazcikov" [The circle of storytellers], Kontinent
11: 129-140.
Doyle, Arthur Conan (Sir)
1981 The complete Sherlock Holmes (London: Seeker & Warburg).
Dragunskij, Viktor
1968 Deniskiny rasskazy [Little Denis' stories] (Moskva: Detskaja
Literatura).
Drazdauskiene, Maria-Liudvika
1981 "On stereotypes in conversation", in: Coulmas (ed.), 1981:55-68.
Eades, Diana
1982 "You gotta know how to talk... : information seeking in South-
East Queensland Aboriginal society", Australian Journal of
Linguistics 2:61-82.
Ervin-Tripp, Susan
1974 "Sociolinguistics", in: Blount (ed.), 1974:268-334.
1976 "Is Sybil there? The structure of some American English
directives", Language in Society 5:25-66.
Facey, Albert B.
1981 A fortunate life (Ringwood, Vic.: Penguin).
Field, Deborah
1988 "Tautologies in Japanese", (unpublished paper, Australian
National University).
Fillmore, Charles
1984 "Remarks on contrastive pragmatics", in: Fisiak (ed.), 119-141.
Fishman, Joshua (ed.)
1968 Readings in the sociology of language (The Hague: Mouton).
Fisiak, Jacek (ed.)
1984 Contrastive linguistics: prospects and problems (Berlin: Mouton).
Fochi, Franco
1966 L' italiano facile [Italian made easy] (Milano: Feltrinelli).
Folsom, Franklin
1977 Kniga 0 jazyke [The language book] (Moskva: Progress).
[1963]
Forgas, Joseph P. (ed.)
1985 Language and social situations (New York: Springer).
468 Bibliography
Fraser, Bruce - Ellen Rintell - Joel Walters
1980 "An approach to conducting research on the acquisition of prag-
matic competence in a second language", in: Larsen-Freeman
(ed.),75-91.
Gage, William W. (ed.)
1974 Language in its social setting (Washington, DC: Anthropological
Society of Washington).
Garton Ash, Timothy
1983 The Polish revolution: Solidarity 1980-82 (London: Jonathan
Cape).
Gazdar, George
1979 Pragmatics: implicature, presupposition and logical form (New
York: Academic Press).
Geertz, Clifford
1976 The religion of Java (Chicago: Chicago University Press).
Gerber, Eleanor
1985 "Rage and obligation: Samoan emotion in conflict", in: White -
Kirkpatrick (eds.), 121-167.
Gibbs, Raymond W., Jr
1985 "Situational conventions and requests", in: Forgas (ed.), 97-110.
Giglioli, Pier Paolo (ed.)
1972 Language and social context (Harmondsworth: Penguin).
Givan, Talmy
1983 "Review" [of Cole (ed.) 1981], Studies in Language 7:151-155.
1984 "The speech-act continuum", in: Chisholm (ed.), 245-254.
1989 Mind, code, and context: essays in pragmatics (Hillsdale, N.J.:
Erlbaum).
Gladwin, Thomas - William Sturtevant (eds.)
1962 Anthropology and human behavior (Washington, DC: Anthropo-
logical Society of Washington).
Goddard, Cliff
1979 "Particles and illocutionary semantics", Papers in Linguistics
12: 185-229.
1985 A grammar of Yankunytjatjara (Alice Springs: Institute for
Aboriginal Development).
1986 "The natural semantics of too", Journal of Pragmatics 10:635-
644.
1989a "Issues in natural semantic metalanguage", Quaderni di Seman-
tica 10.1:51-64 (Round table on semantic primitives, 1).
1989b "The goals and limits of semantic representation", Quaderni
di Semantica 10.2 (Round table on semantic primitives, 2).
Bibliography 469
Goddard, Cliff (ed.)
1987 A basic Pitjantjatjara/Yankunytjatjara to English dictionary
(Alice Springs: Institute for Aboriginal Development).
Godel, Robert (ed.)
1969 A Geneva School reader in linguistics (Bloomington: Indiana
University Press).
Goffman, Erving
1981 Forms of talk (Oxford: Blackwell).
Goldstein, Bernice - Kyoko Tamura
1975 Japan and America: a comparative study in language and culture
(Rutland, Vermont: Tuttle).
Goljavkin, Viktor
1968 Ty prixodi k nam, prixodi [Come to us, come] (Leningrad:
Detskaja Literatura).
Goody, Esther (ed.)
1978 Questions and politeness: strategies in social interaction
(Cambridge: Cambridge University Press).
Gordon, David - George Lakoff
1975 "Conversational postulates", in: Cole - Morgan (eds.), 83-106.
Grandgent, Charles H.
1908 An introduction to Vulgar Latin (Boston: Heath).
Green, Georgia
1975 "How to get people to do things with words: the whimperative
question", in: Cole - Morgan (eds.), 107-141.
Greenberg, Joseph (ed.)
1978 Universals of human language 1-4. (Stanford, Calif.: Stanford
University Press).
Grice, H. Paul
1975 "Logic and conversation", in: Cole - Morgan (eds.), 41-58.
1981 "Presupposition and conversational implicature", in: Cole (ed.),
183-198.
Grochowski, Maciej
1986 Polskie partykuly: skladnia, semantyka, leksykografia [Polish par-
ticles: syntax, semantics, lexicography] (Wroctaw: Ossolineum)
(Prace Instytutu Polskiego PAN 62).
Grodzieriska, Wanda
1970 Swierszczowa muzyka [The crickets' music] (Warszawa: Nasza

Gross, Maurice - Morris Halle - Marcel Schtitzenberger (eds.)
1973 The formal analysis of natural languages (The Hague: Mouton).
Grossman, Vasily
1980 Zizn' i sud' ba [Life and fate] (Paris: l'Homme).
1985 Life and fate (Robert Chandler, trans.) (London: Fontana).
470 Bibliography
Gumperz, John
1972 "Introduction", in: Gumperz - Hymes (eds.), 1-25.
1982 Discourse strategies (Cambridge: Cambridge University Press).
Gumperz, John - Dell Hymes (eds.)
1972 Directions in sociolinguistics (New York: Holt, Rinehart &
Winston).
Haiman, John
1989 "Alienation in grammar", Studies in Language 13:129-170.
Halliday, Michael A.K. - Ruqaiya Hasan
1976 Cohesion in English (London: Longman).
Harasawa, Itsuo
1985 "Japanese tautologies", (unpublished paper, Australian National
University).
Harkins, Jean
1986 "Semantics and the language learner: Warlpiri particles", Journal
of Pragmatics 10:559-573.
1988 English as a 'two-way' language in Alice Springs (M.A. thesis,
Australian National University).
Harris, Max
1962 "Morals and manners", in: Coleman (ed.), 47-67.
Harris, Stephen G.
1984 Culture and learning: tradition and education in north-east
Arnhem Land (Canberra: Australian Institute of Aboriginal
Studies).
Hayakawa, Haruko
1985 The semantics of reduplication in Japanese (M.A. thesis,
Australian National University).
Hibberd, Jack
1974 "Who", "White with wire wheels", in: Four Australian plays
(Ringwood, Vic.: Penguin).
Higa, Masanori
1972 "The use of the imperative mood in post-war Japan", in: Lebra
(ed.), 49-56.
Hijirida, Kyoko - Ho-min Sohn
1986 "Cross-cultural patterns of honorifics and sociolinguistic sensitiv-
ity to honorific variables: evidence from English, Japanese, and
Korean", Papers in Linguistics 19:365-401.
Hill, Deborah
1985 The semantics of some interjectional constructions in
Australian English (B.A.Hons. thesis, Australian National
University).
Bibliography 471
Hirszowicz, Maria
1980 The bureaucratic leviathan: a study in the sociology of commu-
nism (Oxford: Martin Robertson).
Hollos, Marida - William Beeman
1978 "The development of directives among Norwegian and Hungarian
children: an example of communicative style in culture",
Language in Society 7:345-355.
Honna, Nobuyoki - Bates Hoffer
1989 An English dictionary of Japanese ways of thinking (Tokyo:
Yuhikaku).
Hornby, Albert Sydney - E.V. Gatenby - H. Wakefield
1969 The advanced learner's dictionary of current English (2nd
edition) (Oxford: Oxford University Press).
Horne, Donald
1964 The lucky country: Australia in the sixties (Ringwood, Vic.:
Penguin).
House, Juliane - Gabriele Kasper
1981 "Politeness markers in English and German", in: Coulmas (ed.),
157-185.
1962
1974b
1974a
Hubler, Axel
1983 Understatements and hedges in English (Amsterdam: John
Benjamins).
Hudson, Joyce
1985 "Selected speech act verbs in Walmatjari", in: Huttar - Greger-
son (eds.), 1985:63-83.
Hughes, Geoffrey
1988 Words in time: a social history of English vocabulary (Oxford:
Blackwell).
Humboldt, Wilhelm von
1903-18 Gesammelte Schriften 1-7 (A. Leitzmann, ed.) (Berlin: Behrs).
Huttar, George - Kenneth Gregerson (eds.)
1985 Pragmatics in non-Western perspective (Dallas, Texas: Summer
Institute of Linguistics).
Hymes, Dell H.
1961 "On typology of cognitive styles in language (with examples
from Chinookan)", Anthropological Linguistics 3:22-54.
"The ethnography of speaking", in: Gladwin - Sturtevant (eds.),
13-53 [Reprinted in Fishman (ed.), 99-138].
"Anthropology and sociology: an overview", in: Sebeok (ed.),
1445-1475.
Foundations in sociolinguistics: an ethnographic approach
(Philadelphia: University of Pennsylvania Press).
472 Bibliography
Jacobs, Roderick - Peter Rosenbaum (eds.)
1970 Readings in English transformational grammar (Waltham, Mass.:
Ginn).
Jespersen, Otto
1965 A modern English grammar on historical principles 1-7 (London:
[1909-49] Allen & Unwin).
Johnson, Christina
1978 In praise of dogs in Australia (Adelaide: Rigby).
Johnson, Ken
1972 "The vocabulary of race", in: Kochman (ed.), 140-151.
Kaczynski, Mieczyslaw
1964 Gramatyka wloskiego [Grammar of the Italian language]
(Warszawa: Wiedza Powszechna).
Kageyama, Taro - Ikuhiro Tamori
1976 "Japanese whimperatives", Papers in Japanese Linguistics 4:
13-53.
Kalish, Donald - Richard Montague
1964 Logic: techniques of formal reasoning (New York: Harcourt,
Brace).
Karcevski, Serge
1969 "Introduction a de l'interjection", in: Godel (ed.), 196-212.
[1941]
Katriel, Tamar
1986 Talking straight: dugri speech in Israeli Sabra culture (Cam-
bridge: Cambridge University Press).
Keen, Ian D.
1978 One ceremony, one song: an economy of religious knowledge
among the Yolngu of north-east Arnhem Land (Ph.D. thesis,
Australian National University).
Kemme, Hans-Martin
1979 Ja, denn, doch usw, Die Modalpartikeln im Deutschen (Miinchen:
Goethe-Institut).
Kendon, Adam - Richard Harris - Mary Ritchie Key (eds.)
1975 Organization of behavior in face-to-face interaction (The Hague:
Mouton).
Kiefer, Ferenc - Nicholas Ruwet (eds.)
1973 Generative grammar in Europe (Dordrecht: Reidel).
Kinchin-Smith, F. - T.W. Melluish
1966 Teach yourself Greek (London: The English Universities Press).
King, Jonathan
1978 Waltzing materialism (Sydney: Harper & Row).
Bibliography 473
Kochman, Thomas
1972 "Toward an ethnography of black American speech behavior", in:
Kochman (ed.), 241-264.
1981 Black and white styles in conflict (Chicago: University of Chicago
Press).
Kochman, Thomas (ed.)
1972 Rappin' and stylin' out: communication in urban black America
(Urbana: University of Illinois Press).
Kolankiewicz, George - Paul Lewis
1988 Poland: politics, economics, and society (London: Pinter).
Konig, Ekkehard
1977 "Temporal and non-temporal uses of 'noch' and 'schon' in
German", Linguistics and Philosophy 1:173-198.
Kucera, Henry - W. Nelson Francis
1968 Computational analysis of present-day American English (Provi-
dence, R.I.: Brown University Press).
Kurokawa, Shozo
1972 "Japanese terms of address: some usages of the first and second
person pronouns", Papers in Japanese Linguistics 1:228-238.
Labov, William
1972 "Rules for ritual insults", in: Sudnow (ed.), 120-169 [Reprinted in
Kochman (ed.), 265-314, and in Labov, Language in the inner
city, 297-353 (Philadelphia: University of Pennsylvania Press)].
Lakoff, George
1986 Women, fire, and dangerous things: what categories reveal about
the mind (Chicago: University of Chicago Press).
Lakoff, Robin
1972 "Language in context", Language 48:907-927.
Lanham, Betty B.
1986 "Ethics and moral precepts taught in schools of Japan and the
United States", in: Lebra - Lebra (eds.), 280-296.
Larsen-Freeman, Diane (ed.)
1980 Discourse analysis in second language research (Rowley, Mass.:
Newbury House).
LDOCE
1978
LDOTEL
1984
Longman dictionary of contemporary English (Paul Procter, ed.)
(London: Longman).
Longman Dictionary of the English Language (Heather Gay, Brian
O'Kill, Katherine Seed, Janet Whitcut, eds.) (London: Longman).
474 Bibliography
Lebra, Takie S.
1974 "Reciprocity and the asymmetric principle: an analytic reappraisal
of the Japanese concept of On", in: Lebra - Lebra (eds.), 192-
207.
1976 Japanese patterns of behavior (Honolulu: University Press of
Hawaii).
Lebra, Takie S. - William P. Lebra (eds.)
1974 Japanese culture and behavior: selected readings (Honolulu:
University Press of Hawaii).
1986 Japanese culture and behavior: selected readings (revised edition)
(Honolulu: University Press of Hawaii).
Lebra, William P. (ed.)
1972 Transcultural research in mental health (Honolulu: University
Press of Hawaii).
Lee, David
1987 "The semantics of just", Journal of Pragmatics 11 :377-398.
Leech, Geoffrey
1983 Principles of pragmatics (London: Longman).
Leibniz, Gottfried Wilhelm
1903 Opuscules et fragments inedits de Leibniz (Louis Couturat, ed.)
(Paris) [Reprinted 1961, Hildesheim: Georg DIms].
1949 New essays concerning human understanding (La Salle, Ill.: Open
[1704] Court).
Lepschy, Anna - Giulio Lepschy
1984 La lingua italiana [The Italian language] (Milano: Bompiani).
Levenston, Edward A.
1970 English for Israelis (Tel Aviv: Israeli Universities Press).
Levinson, Stephen
1983 Pragmatics (Cambridge: Cambridge University Press).
Levy, Robert
1973 Tahitians: mind and experience in the Society Islands (Chicago:
University of Chicago Press).
Lewis, Clive Staples
1946 The Screwtape letters (London: Geoffrey Bles).
Locke, John
1959 [1690] An essay concerning human understanding 1-2. (A.C. Fraser,
ed.) (New York: Dover).
Luo, Yongxian
1988 "An account of Chinese nominal tautologies", (unpublished paper,
Australian National University).
Lutz, Catherine
1982 "The domain of emotion words on Ifaluk", American Ethnologist
9:113-128.
Bibliography 475
1985 "Ethnopsychology compared to what? Explaining behavior and
consciousness among the Ifaluk", in: White - Kirkpatrick (eds.),
35-79.
1986 "Emotion, thought, and estrangement: emotion as a cultural
category", Cultural Anthropology 1: 287-309.
1988 Unnatural emotions: everyday sentiments on a Micronesian
atoll and their challenge to Western theory (Chicago: University
of Chicago Press).
Malamud, Bernard
1958 The magic barrel (New York: Farrar, Straus).
1971 The tenants (New York: Farrar, Straus).
Manzoni, Alessandro
1914 The betrothed (Bohn's Libraries, trans.) (London: Bell).
1968 The betrothed (Archibald Colquhoun, trans.) (London: Dent).
1972 I promessi sposi [The betrothed] (Verona: Mondadori).
Marsella, Anthony - George DeVos - Francis Hsu (eds.)
1985 Culture and self: Asian and Western perspectives (New York:
Tavistock).
Matisoff, James
1979 Blessings, curses, hopes, and fears: psycho-ostensive expressions
in Yiddish (Philadelphia: Institute for the Study of Human Issues).
Matsumoto, Yoshiko
1988 "Reexamination of the universality of face: politeness phenomena
in Japanese", Journal of Pragmatics 12:403-426.
McGregor, Craig
1980 The Australian people (Sydney: Hodder & Stoughton).
Mel' cuk, Igor
in press Cours de morphologie generale. (Montreal: Presses Universi-
taires).
Mezzrow, Milton - Bernard Wolfe
1969 Really the blues (Garden City, N.Y.: Doubleday).
Mickiewicz, Adam
1955 "Dziady, z ~ III" [Grandfathers, part 3] (Utwory zebrane
[1832] [Collected works]) (Warszawa: Czytelnik).
Miller, Roy
1967 The Japanese language (Chicago: University of Chicago Press).
Minami, Hiroshi
1953 Nihonjin no shinri [The psychology of the Japanese] (Tokyo:
Iwanami).
Mitchell-Kernan, Claudia
1971 Language behavior in a black urban community (Berkeley:
University of California Press) (Monographs of the Language
Behavior Research Laboratory, 2).
476 Bibliography
1972 "Signifying and marking: two Afro-American speech acts", in:
Gumperz - Hymes (eds.), 161-179.
Mittwoch, Anita
1977 "How to refer to one's own words: speech act modifying adverbi-
als and the performative analysis", Journal of Linguistics 13:
177-189.
Mizutani, Osamu - Nobuko Mizutani
1987 How to be polite in Japanese (Tokyo: Japan Times).
Moiseev, Aleksandr
1978 "Casticy uze i jelce v sovremennom russkom jazyke" [The
particles uze and jesce in contemporary Russian], Slavia Orien-
talis 27:357-360.
Monahan, Barbara
1983 A dictionary of Russian gesture (Ann Arbor, Mich.: Hermitage).
Montgomery, Lucy
1980 Anne of Ingleside (London: Angus & Robertson).
Moravcsik, Edith
1978 "Reduplicative constructions", in: Greenberg (ed.), 3:297-334.
Morgan, Jerry
1978 "Two types of convention in indirect speech acts", in: Cole (ed.),
261-280.
Morimoto, Junko
1985 "Some Japanese speech act verbs", (unpublished paper, Australian
National University).
Morris, Charles W.
1938 "Foundations of the theory of signs", in: Neurath - Carnap -
Morris (eds.), 77-137.
Morris, Desmond - Peter Collett - Peter Marsh - Marie O'Shaughnessy
1979 Gestures: their origins and distribution (London: Jonathan Cape).
Myerhoff, Barbara
1978 Number our days (New York: Dutton).
Nabokov, Vladimir
1983 Pnin (Ann Arbor: Ardis).
Nakane, Chie
1970 Japanese society (London: Weidenfeld & Nicolson).
1972 Human relations in Japan: summary translation of "Tateshakai
no ningen kankei" [Personal relations in a vertical society]
(Tokyo: Ministry of Foreign Affairs).
Nesbitt, Edith
1980 The enchanted castle (Harmondsworth: Penguin).
Neurath, Otto - Rudolf Carnap - C.W. Morris (eds.)
1938 International encyclopaedia of unified science (Chicago: Chicago
University Press).
Bibliography 477
Nevile, Ann
1981 A comparison of selected speech acts in Japanese and English
(B.A.Hons. thesis, Australian National University).
Newmeyer, Frederick
1980 Linguistic theory in America: the first quarter-century of transfor-
mational generative grammar (New York: Academic Press).
Nikolaeva, Tat'jana M.
1985 Funkcji castic v vyskazyvanii: na materiale slavjanskix jazykov
[Functions of particles in the utterance: on materials from Slavic
languages] (Moskva: Nauka).
Nuchelmans, Gabriel
1973 Theories of the proposition: ancient and medieval conceptions
of the bearers of truth and reality (Amsterdam: North-Holland).
Ochs Keenan, Elinor
1976 "The universality of conversational postulates", Language in
Society 5:67-80.
OED
1933 The Oxford English dictionary 1-12. (Oxford: Clarendon Press).
OEDS
1972-86 A supplement to the Oxford English dictionary 1-4. (R.W. Burch-
field, ed.) (Oxford: Clarendon Press).
O'Grady, John
1965 Aussie English: an explanation of the Australian idiom (Sydney:
Ure Smith).
Okudzava, Bulat
1976 Arbat, moj Arbat: stixi i pesni [Arbat, my Arbat: poems and songs]
(Moskva: Sovetskij Pisatel').
O'Neill, Eugene
1958 "Welded", in: Three plays by Eugene O'Neill (London: Jonathan
Cape).
1960 Anna Christie (Harmondsworth: Penguin).
Dring, Elliott
1981 Israeli humor: the content and structure of the 'chizbat' of the
Palmah (Albany: SUNY Press).
Paduceva, Elena
1985 Vyskazyvanie i ego sootnesennost' s dejstvitel' nost'ju [The utter-
ance and its relation to reality] (Moskva: Nauka).
Palakomkul, Angkab
1975 "A socio-linguistic study of pronominal usage in spoken Bangkok
Thai", International Journal of the Sociology of Language 5:
11-41.
Pascal, Fania
1981 "Wittgenstein: a personal memoir", in: Rhees (ed.), 26-62.
478 Bibliography
1982
1973
"Doch - eine Studie zur Syntax und zur kommunikativen
Funktion einer Partikel", Deutsche Sprache 1975: 222-242.
Rayner, Clare
1978 The body book (Ultimo, NSW: Whizzard Publications).
Reisman, Karl
1974 "Noise and order", in: Gage (ed.), 50-61.
Renwick, George W.
1980 InterAct: guidelines for Australians and North Americans
(Yarmouth, Maine: Intercultural Press).
Rhees, Rush (ed.)
1981 Ludwig Wittgenstein: personal recollections (Oxford: Blackwell).
Rosaldo, Michelle
1980 Knowledge and passion: Ilongot notions of self and social life
(Cambridge: Cambridge University Press).
"The things we do with words: Ilongot speech acts and speech
act theory in philosophy", Language in Society 11:203-237.
Ross, John R.
1970 "On declarative sentences", in: Jacobs - Rosenbaum (eds.), 222-
272.
"Slifting" [sentence lifting], in: Gross - Halle - Schutzenberger
(eds.), 133-169.
Pasicki, Adam
1976 "lui, jeszcze, and their English equivalents", Papers and Studies
in Contrastive Linguistics (Poznan;) 5: 103-110.
Paterson, Katherine
1978 The great Gillie Hopkins (Harmondsworth: Penguin).
Pomerantz, Anita
1978 "Compliment responses: notes on the co-operation of multiple
constraints", in: Schenkein (ed.), 79-112.
Pride, Janet B. (ed.)
1985 Cross-cultural encounters: communication and mis-communica-
tion (Melbourne: River Seine).
Psathas, George
1979 Everyday language: studies in ethnomethodology (New York:
Irvington).
Quirk, Randolph - Sidney Greenbaum - Geoffrey Leech - Jan Svartvik
1972 A grammar of contemporary English (London: Longman).
Raphael, Frederic
1977 The glittering prizes (Harmondsworth: Penguin).
Rath, Rainer
1975
Rosten, Leo
1968 The joys of Yiddish (New York: McGraw-Hill).
Bibliography 479
Russell, Bertrand
1962 An inquiry into meaning and truth (Harmondsworth: Penguin).
Rymkiewicz, Jaroslaw Marek
1984 Rozmowy polskie latem 1983 [Polish conversations, summer 1983]
(Paris: Instytut Literacki).
Sacks, Harvey - Emmanuel Schegloff - Gail Jefferson
1978 "A simplest systematics for the organization of turn taking for
conversation", in: Schenkein (ed.), 7-55.
Sadock, Jerrold
1970 "Whimperatives", in: Sadock - Vanek (eds.), 223-238.
1974 Toward a linguistic theory of speech acts (New York: Academic
Press).
1981 "Almost", in: Cole (ed.), 257-271.
Sadock, Jerrold - Anthony Vanek (eds.)
1970 Studies presented to R.B. Lees by his students (Edmonton:
Linguistic Research Inc.).
Sansom, Basil
1980 The camp at Wallaby Cross: Aboriginal fringe dwellers in Darwin
(Canberra: Australian Institute of Aboriginal Studies).
Sapir, Edward
1921 Language: an introduction to the study of speech (New York:
Harcourt, Brace).
1949 "Gradation: a study in semantics", in: Selected writings of Edward
Sapir in language, culture, and personality, 122-149 (David
Mandelbaum, ed.) (Berkeley: University of California Press).
Schachter, Josef
1973 Prolegomena to a critical grammar (Dordrecht: Reidel) [First
[1935] published 1935, Vienna: Julius Springer].
Schenkein, Jim (ed.)
1978 Studies in the organization of conversational interaction (New
York: Academic Press).
Schieffelin, Bambi B. - Eleanor Dchs (eds.)
1986 Language socialization across cultures (Cambridge: Cambridge
University Press) (Studies in the social and cultural foundations
of language, 3).
Schiffrin, Deborah
1984 "Jewish argument as sociability", Language in Society 13:311-
335.
Schreiber, Paul
1972 "Style disjuncts and the performative analysis", Linguistic Inquiry
3:321-347.
480 Bibliography
Schrett, Jozef
1984 "Dolina gorami" [A valley between mountains], Kultura
436:5-18.
Schuchardt, Hugo
1972 "On sound laws: against the Neogrammarians", in Vennemann -
Wilbur (eds.), 39-72. [First published 1885.]
1975
1979
Searle, John
1969
"Namakke, sunmakke, kormakke: three types of Cuna speech
event", in: Bauman - Sherzer (eds.), 263-282.
Shetter, William Z.
1966 "The meaning of German noch", Language 42:42-66.
Shopen, Timothy
1974 "Some contributions from grammar to the theory of style",
College English 35:775-798.
Shweder, Richard - Robert LeVine (eds.)
1984 Culture theory: essays on mind, self, and emotion (Cambridge:
Cambridge University Press).
Siatkowski, Janusz
1977 Studia Streszczenia prac doktorskich II [Linguistic
studies: abstracts of doctoral dissertations II] (Wroctaw:
Ossolineum) (Prace PAN 88).
SJP
1958-68 Slownik polskiego (Witold Doroszewski, ed.) (Warszawa:
Panstwowe Wydawnictwo Naukowe).
Skinner, Quentin
1970 "Conventions and the understanding of speech acts", The Philo-
sophical Quarterly 20: 118-138.
Skorupka, Stanislaw
1974 Slownik frazeologiczny polskiego [Phraseological
dictionary of the Polish language] 1-2. (2nd edition) (Warszawa:
Wiedza Powszechna).
Speech acts: an essay in the philosophy of language (Cambridge:
Cambridge University Press).
"Indirect speech acts", in: Cole - Morgan (eds.), 59-82.
Expression and meaning: studies in the theory of speech acts
(Cambridge: Cambridge University Press).
Sebeok, Thomas A. (ed.)
1974 Current trends in linguistics XII: Linguistics and adjacent arts
and sciences (The Hague: Mouton).
Sekiguchi, Isugio
1977 "Was heisst doch?", in Weydt (ed.), 3-9.
Sherzer, Joel
1974
SRJa
1957-61
Bibliography 481
Smith, Hedrick
1976 The Russians (London: Sphere Books).
Smith, Robert
1983 Japanese society: tradition, self, and the social order (Cambridge:
Cambridge University Press).
Sohn, Ho-min
1983 "Intercultural communication in cognitive values: Americans and
Koreans", Language and Linguistics (Seoul) 9:93-136.
Solomon, Robert
1984 "Getting angry: The Jamesian theory of emotion in anthropology",
in: Shweder - LeVine (eds.), 238-254.
Sperber, Dan - Deirdre Wilson
1981 "Irony and the use-mention distinction", in: Cole (ed.), 295-318.
Spitzer, Leo
1928 Stilstudien 1-2. (MUnchen: Hueber).
Spradley, James P. - Michael A. Rynkiewich
1975 The Nacirema: readings on American culture (Boston: Little,
Brown).
Slovar' russkogo jazyka [Dictionary of the Russian language]
1-4. (Moskva: Gosudarstvennoe izdatel' stvo inostrannyx i
nacional 'nyx slovarej).
SSRLJa
1950-65 Slovar' sovremennogo russkogo literaturnogo jazyka [Dictionary
of contemporary literary Russian] 1-17. (Moskva: Izdatel' stvo
Akademii Nauk SSSR).
Stevenson, Burton
1946 Stevenson's book of quotations (London: Cassell).
1949 Stevenson's book of proverbs, maxims and familiar phrases
(London: Routledge & Kegan Paul).
Sudnow, David (ed.)
1972 Studies in social interaction (New York: Free Press).
Sumerkin, Alexander (ed.)
1982 Russica 1981: literaturnyj sbornik [Russica 1981: a literary
collection] (New York: Russica).
Suslov, 11 'ja
1982 "Inga i Verocka" [Inga and Verochka], in: Sumerkin (ed.), 183-
190.
Suzuki, Takao
1986 "Language and behavior in Japan: the conceptualization of per-
sonal relations", in: Lebra - Lebra (eds.), 142-157.
482 Bibliography
Szelburg-Zarembina, Ewa
1970 Idzie niebo ciemn{l noc{l ... [The sky is walking on a dark night. .. ]
(Warszawa: Nasza
Szwed, John F. (ed.)
1970 Black America (New York: Basic Books).
Tannen, Deborah
1981a "Indirectness in discourse: ethnicity as conversational style",
Discourse Processes 4:221-238.
1981b "New York Jewish conversational style", International Journal
of the Sociology of Language 30:133-149.
1984 Conversational style: analyzing talk among friends (Norwood,
N.J.: Ablex).
1986 That's not what I meant! How conversational style makes or
breaks relationships (New York: Ballantine).
Tannen, Deborah - Piale Oztek
1977 '" Health to our mouths': formulaic expressions in Turkish and
Greek", Berkeley Linguistics Society, Proceedings 3: 516-534
[Reprinted in Coulmas (ed.), 37-54].
Taylor, Brian A.
1976 "Towards a sociolinguistic analysis of 'swearing' and the
language of abuse in Australian English", in: Clyne (ed.), 43-62.
Tolstoy, Nikolai
1983 The Tolstoys: twenty-four generations of Russian history 1353-
1983 (London: Harnish Hamilton).
Traugott, Elizabeth C. - John Waterhouse
1969 '" Already' and 'yet': a suppletive set of aspect-markers?",
Journal of Linguistics 5:287-304.
Triandaphyllidis, Manolis
1975 Petite grammaire du grec moderne (F. Duisit, O. Merlier, trans.)
(Thessaloniki).
Triandis, Harry C. - Leigh M. Triandis
1960 "Race, social class, religion, and nationality as determinants of
social disclosure", Journal of Abnormal and Social Psychology
61: 110-118.
Triandis, Harry C. - Vasso Vassiliou
1972 "A comparative analysis of subjective culture", in: Triandis (ed.),
299-335.
Triandis, Harry C. (ed.)
1972 The analysis of subjective culture (New York: Wiley).
Trilling, Lionel
1972 Sincerity and authenticity (London: Oxford University Press).
Ueda, Keiko
Bibliography 483
1974 "Sixteen ways to avoid saying 'no' in Japan", in: Condon - Saito
(eds.), 184-192.
Universite de Paris-7
1986 Les particules enonciatives en russe contemporain (Paris:
Departement de Recherches Linguistiques (DRL), Laboratoire de
Linguistique Formelle) (ATP Nouvelles recherches sur Ie langage,
Collection ERA 642).
Urmson, James Opie
1963 "Parenthetical verbs", in: Caton (ed.), 220-240.
Vennemann, Theo - Terence H. Wilbur (eds.)
1972 Schuchardt, the Neogrammarians, and the transformational
theory of phonological change (Frankfurt: Athenaum).
Verschueren, Jef
1985 What people say they do with words: prolegomena to an empirical
conceptual approach to linguistic action (Norwood, N.J.: Ablex).
Vossler, Karl
1904 Positivismus und Idealismus in der Sprachwissenschaft
(Heidelberg).
1925 Geist und Kultur in der Sprache (Munchen).
Wannan, Bill
1963 Tell' em I died game (Melbourne: Lansdowne Press).
Ward, Russel
1958 The Australian legend (Melbourne: Oxford University Press).
Weber, Max
1968 [1930] The Protestant ethic and the spirit of capitalism (Talcott
Parsons, trans.) (London: Allen & Unwin).
Webster
1973 The international Webster new encyclopaedic dictionary (New
York: Tabor House).
1977 Webster's new world dictionary (Sydney: Fontana/Collins).
Weydt, Harald
1969 Abtonungspartikel, Die deutschen Modalworter und ihre
franzosischen Entsprechungen (Bad Homburg: Gehlen).
Weydt, ~ r l d (ed.)
1977 Aspekte der Modalpartikeln: Studien zur deutschen Abtonung
(Ttibingen: Niemeyer).
Weydt, Harald - Theo Harden - Elke Hentschel - Dietmar RosIer (eds.)
1983 Kleine deutsche Partikellehre (Stuttgart: Klett).
White, Geoffrey M. - John Kirkpatrick (eds.)
1985 Person, self, and experience: exploring Pacific ethnopsychologies
(Berkeley: University of California Press).
484 Bibliography
1974
1987
1972
1984
1990
1988
1989a
1989b
1986b
1986c
1985c
1986a
1985b
1985a
1980
1976
1977
1973
Whorf, Benjamin Lee
1956 Language, thought, and reality (John Carroll, ed.) (New York:
Wiley).
Wierzbicka, Anna
1971 "The deep or semantic structure of the comparative", Linguisti-
sche Berichte 16:39-45.
Semantic primitives (Frankfurt: Athenaum) (Linguistische
Forschungen, 22).
"The semantic structure of words for emotions", in: Jakobson -
van Schooneveld - Worth (eds), Slavic poetics, 499-505 (The
Hague: Mouton).
"The semantics of direct and indirect discourse", Papers in
Linguistics 7:267-307.
"Particles and linguistic relativity", International Review of
Slavic Linguistics 1:327-367.
"The ignorative: the semantics of speech acts", International
Review of Slavic Linguistics 2:251-312.
Lingua mentalis: The semantics of natural language (Sydney/
New York: Academic Press).
"Diminutives and depreciatives: semantic representation for
derivational categories", Quaderni di Semantica 5: 123-130.
"A semantic metalanguage for a cross-cultural comparison of
speech acts and speech genres", Language in Society 14:491-514.
"Different cultures, different languages, different speech acts:
Polish vs. English", Journal of Pragmatics 9:145-178.
Lexicography and conceptual analysis (Ann Arbor: Karoma).
"A semantic metalanguage for the description and comparison of
illocutionary meanings", Journal of Pragmatics 10:67-107.
"Does language reflect culture? Evidence from Australian
English", Language in Society 15:349-373.
"Human emotions: universal or culture-specific?", American
Anthropologist 88:584-594.
English speech act verbs: A semantic dictionary (Sydney:
Academic Press).
The semantics of grammar (Amsterdam: John Benjamins).
"Semantic primitives and lexical universals", Quaderni di
Semantica 10.1:103-121 (Round table on semantic primitives, 1).
"Semantic primitives: the expanding set", Quaderni di Semantica
10.2 (Round table on semantic primitives, 2).
"Antitotalitarian language in Poland: some mechanisms of linguis-
tic self-defence", Language in Society 19: 1-59.
Bibliography 485
in press a "Lexical universals and universals of grammar", in Michel
Kefer - lohan van der Auwera (eds.) Meaning and grammar.
(Berlin: de Gruyter).
in press b "Semantic complexity", to appear in Theoretical Linguistics 17,
1990 [special issue on "Complexity"].
to appear Semantics, culture and cognition: universal human concepts in
culture-specific configurations.
Wierzbicka, Anna (ed.)
1972 Semantyka i slownik [Semantics and lexicon] (Wroclaw:
Ossolineum).
Wilkes, Gerald A.
1978 A dictionary of Australian colloquialisms (Sydney: Fontanal
Collins).
Wilkins, David P.
1984 "Nominal reduplication in Mparntwe Arrernte", Language in
Central Australia 1: 16-22.
1986 "Particle/clitics for criticism and complaint in Mparntwe Arrernte
(Aranda)", Journal of Pragmatics 10:575-596.
Williams, Tennessee
1961 The night of the iguana (London: Secker & Warburg).
Williamson, David
1973 Don's party (Collected plays, 1) (Sydney: Currency Press).
1974 The coming of Stork, Jugglers three, What if you died tomorrow:
three plays (Sydney: Currency Press).
Wittgenstein, Ludwig
1953 Philosophical investigations (G.E.M. Anscombe, trans.) (Oxford:
Blackwell).
Zapolska, Gabriela
1978 Moralnose pani Dulskiej [Mrs. Dulska's morality] (Wroc}aw:
Ossolineum).
Zasorina, L.N. (ed.)
1977 Castotnyj slovar' russkogo jazyka [Frequency dictionary of the
Russian language] (Moskva: Russkij Jazyk).
Zipf, George K.
1949 Human behavior and the principle of least effort: an introduction
to human ecology (Facsimile edition, 1965) (New York: Hafner).
Subject and name index
Abby, Dear 143-144
Abelard, Peter 198, 253
Aboriginal (Australian) culture 63,
102-103, 159-60
Abrahams, Roger 63, 69, 103, 150,
183-184,277
abuse, verbal 35, 40, 232-235, 249-251
address forms 48,51,56-59,86, 106-
107, 111-112, 122
advice 31-32
agreement 140
Ali, Mohammed 68
Ameka, Felix 69, 131,342,344,459
American culture, see English,
American
AND (Australian National Dictionary)
165-168,170-178,181-183
Anglo-American culture, see English,
American
Anglo-Saxon culture 26, 47-60, 64-65,
69,72-77, 103-104,276-277,279,
446-448, 454
animals, calls to 292-293
apologies 86, 126, 152, 156-157
Arensberg, Conrad 91
argument 68-69
Aristotle 161
Arnold, Matthew 116-117
Ascham, Roger 104
assertiveness, see self-assertion
Atlas, Jay 396
augmentatives 1
Austin, John L. 197, 240
Australian culture, see English,
Australian; Aboriginal culture
autonomy 37,49,52,76, 80-82, 86,91,
95, 99, 276
Bach, Kent 198,200-201,209,253,
449
Bacon, Roger 198, 253
Baker, Sidney 3, 177
Bakhtin, see Baxtin
Bally, Charles 393,445,451
Bankowski, Andrzej 367
baptism 151-152
Bamlund, Dean 12, 53, 82, 86, 93,
103, 106, 110, 113-114, 129
Bartlett, John 401, 415
Barzini, Luigi 280-282
Bates, Elizabeth 77
Baxtin, Mixail 149, 197,457
Beeman, William 62, 77
Benedict, Ruth 57,71,129,176
Black (American) culture, see
English, Black
blessings 4, 122, 150
Blum-Kulka, Shoshana 77, 89-92,
96-98, 138, 141
Blyton, Enid 415
boasting 84-85, 119, 198
Bogardus, Emory 108
Boguslavskij, Igor' 342
Bogus.fawski, Andrzej 7-8, 164,217,
344, 397, 412
Bohn's Library 438
Bolinger, Dwight 248, 268-269, 271
Bowles, Colin 3
Braun, Friederike 457
Brown, Penelope 30, 67, 108-109,
279,444-445,454
Brown, Roger 47, 58
Brzechwa, Jan 306-307
BUhler, Karl 288
Butwin, Frances 123
488 Subject and name index
Buzo, Alexander 27,29,35,39,42-
44,46, 171
calling 292-293, 300-301
Cattell, Ray 226
change, language 169-170, 173, 177,
189-190
Chekhov, Anton 300-301
Chinese 144, 423-428, 435, 454
Chomsky, Avram Noam 14, 164,
283
circularity 6, 161
Clancy, Patricia 113-114
Clark, Herbert 25
closeness, see intimacy
codes 67
cognitive style 282-283
Cole, Peter 16, 392
Colquhoun, Archibald 438
commands 30, 39, 61, 150-151, 199,
203, 261, 295; see also orders;
imperatives
communicative style 454-455
complaints 60-61, 64-65, 181
complementarism 16-17
complements 221, 238-239
compliments, responses to 136-147
Comrie, Bernard 77,202-205
concessive tautologies 423-426
Confucius 427
conjunctions 240-243, 381
conversation 18, 21, 59-64, 80-82,
116,118,120,131-148,149-151,
202,219-235,253-254,449,451
Conway, Ronald 183
Cooke, Joseph 12
cordiality 27, 50-56, 70, 86
Coulmas, Florian 86, 156, 163
courtesy, see politeness
Cowie, Anthony P. 398,401,440
Crawford, Raymond M. 177
criticism 213-214, 230
cultural differences 2-4, 25-65; see
also values
cultural kinds 17-18
cultural style 276-284; see also
values; identity
Cuna 150
curses 4, 122-123,213,249-251,
338; see also swearing
Dakota 302
Dal', Vladimir 305
Danet, Brenda 77,89-90, 138, 141
Danish 307,313-314,338
Darwin, Charles 312-313, 315-316
Davies, Norman 58, 112, 192-193
Davison, Alice 207
Dawes, Arthur 166
de Tocqueville, Alexis 90
Deakin, Greg 163,247-248
deference, see politeness; honorifics;
address forms
definitions, see metalanguage
DeVos, George 154-155
diminutives 1, 51-53, 55-56, 255,
280
Dineen, Anne 307
directives, see orders; commands;
requests; imperatives
directness 44, 60, 63-64, 70-71, 88-
104, 186-187,454
discourse 258-260
disgust 302-317, 338
dissimulation 63-64, 100-104
distance 37, 47-48, 108-111; see
also calling; intimacy
Dixon, Robert M.W. 273
Djilas, Milovan 193
Doherty, Monika 367
Doi, Takeo 74, 76, 113, 155
Doroszewski, Witold 49
Dostoevsky, Fedor 280, 312
Dovlatov, Sergej 327
downgrade 140-141
Doyle, Arthur Conan 419
Dragunskij, Viktor 327, 336
Drazdauskiene, Maria-Liudvika 43,
45, 116
Dumas, Alexandre 285
Eades, Diana 63,69, 102-103
emotions 53-57,87-88,119,121-
129,244-245, 266-269, 271,275,
279-280, 302-326, 331, 339
emphasis 258; see also intensifica-
tion; intonation
English, see specific subject head-
ings; Black (American) 2-3, 21,
26,63,68,71,73,78-79,82-87,
97,119,123-126,150,165,183-
185,277,454; American 2, 26,
42, 48, 67, 71, 72-82, 85-87, 98-
99,103-108,110,113-121,124-
126, 128, 136, 144, 248, 339,
371, 454-455; Australian 2-3, 20-
22, 26, 42, 48, 55-56, 91, 111-
112, 116, 136, 165-183, 185, 187-
188,249,277,303-304,454,459-
460; Scottish 180
equivalence, semantic 12-14, 152,
286
Ervin-Tripp, Susan 30, 58, 60
ethnocentrism 21, 25-26, 67-68, 71,
131, 150,454
ethnography, of speaking 67, 85,
122, 126, 149-151
exclamations 31, 45-46, 222,
232-235, 249-251; see also
interjections
explication, see metalanguage
face 67, 454
Facey, Albert 357
fatalism 431-433
Field, Deborah 396, 428-430
Subject and name index 489
Fillmore, Charles 220
Finnish 293
first person, in definitions 162-165
fixed expressions 56, 245-247
Fochi, Franco 271
folk labels 150-160
Folsom, Franklin 334-335
formality 48,57, 104, 111-113,358;
see also politeness; distance
Forster, E.M. 395
Francis, W. Nelson 339
Fraser, Bruce 26, 152
French 1, 13-14,71, 103,270,392-
394,445
future 236, 431-433
fuzziness, see indeterminacy
Garton Ash, Timothy 189-190, 192
Gazdar, Gerald 451
Geertz, Clifford 63,87,92, 100-101,
128-129
gender, grammatical 1; see also sex
differences
generative semantics 15-16, 18, 197-
198, 253
genres, speech 149-165, 183-196
Gerber, Eleanor 339
German 13-14,60-61, 103, 286, 302,
304,310,313-315,338,342,367,
370, 375-376, 392-393, 445
gesture 446
Gherson, Rimona 77, 89-90, 138,
141
Gibbs, Raymond 25
Gilman, Albert 47, 58
Givan, Talmy 203-205, 397
Goddard, Cliff 7-8,69,243,293,
344, 346
Goffman, Erving 150, 289-290
Goldstein, Bernice 120-121, 127-128
Goljavkin, Viktor 311
103,276
490 Subject and name index
Gordon, David 32, 59, 62, 213
grammaticality 209-210, 232, 253,
449-452
Grandgent, Charles 255
Greek 95-99, 128,240,315-316,
457-458
Green, Georgia 32-33, 211
greetings 3, 132, 246
Grice, H. Paul 21,23,59,62,67,
366, 391-392, 396-397, 398-400,
403, 449-452, 454
Grochowski, Maciej 344, 367, 377
Grodzienska, Wanda 290, 320
Grossman, Vasily 394
Gumperz, John 62-63, 150
Haiman, John 288, 302, 304
Halliday, Michael 240, 242
Hamlet 118
Harasawa, Itsuo 396
Hardy, Frank 166
Harkins, Jean 69,267,277,307,344
harmony 92, 104, 109-110, 113-115,
118-119
Harnish, Robert 198, 200-201, 209,
253, 449
Harris, Max 169
Harris, Stephen 103, 158
Hasan, Ruqaiya 240, 242
Hawke, Robert 2
Hayakawa, Haruko 270
Hebrew 21, 71, 89-93, 119, 165,
185-188,454
hedges 43-44
Henschke, Sharon 135
Hibberd, Jack 27-29,35,213
Higa, Masanori 30
Hijirida, Kyoko 48,69, 105-107,
111
Hill, Deborah 103, 459
Hirszowicz, Maria 192
Hoffer, Bates 68, 93, 127, 147
Hollos, Marida 62, 77
Honna, Nobuyoki 68, 93, 127, 147
honorifics 12; see also address
forms; politeness
Hornby, Albert 161
Horne, Donald 167, 181-182
House, Juliane 60
HUbler, Axel 276
Hudson, Joyce 159-160
Hughes, Geoffrey 103
Humboldt, Wilhelm von 282
Hungarian 77
Hymes, Dell 67, 149-150, 165,258,
283, 341
identity, cultural 129-130
illocutionary force 17, 22, 60-61,
150, 158, 161-165, 197-254,288-
289, 342
illocutionary grammar 258-260, 276-
280
imperatives 27, 30-32, 36-37, 51-53,
60,76-78, 89, 122, 204-205, 227-
229,231,243,258-260,263,451
implicature 207, 228, 255, 261, 391,
397-403; see also logic
indefinables, see metalanguage
indeterminacy 197-210,212,224,
253-254, 351, 389
indirectness, see directness
infinitive 36-37
informality, see formality; intimacy
insults 151, 183-185, 187,232-233;
see also abuse
intensification 168-170, 256-258,
264, 268-270, 278
intentions 164, 198
intercultural understanding 9, 21, 24,
64-65, 69-70, 129-130, 144, 147-
148, 210, 455
interjections 23, 45-46, 243-245,
285-339; cognitive 326-336;
emotive 302-326; to animals 292-
293; volitive 292-301
interrogative, see questions
intimacy 47-48, 70, 86, 104-111; see
also distance
intonation 247-248
invitations 29, 132, 174, 211-212,
215-216, 228
Israeli Hebrew, see Hebrew
Italian 22, 51, 71, 77, 103,255-284,
435-436,438-439,446,454
Japanese 17, 21,48, 68, 70-71, 145-
147,270,454; culture 2,53-54,
57,72-83,85-88,93-95, 103,
105-107, 112-114, 120-121,
126-128,129-130,144,176;
imperatives 30-31, 88, 258-259;
pronouns 12-14; speech acts 152-
158; tautologies 396, 428-431
Javanese 48, 57, 63-64, 92, 100-104,
128-129
Jefferson, Gail 254
Jespersen, Otto 277, 279,401
Jewish culture 2-4, 68-69, 71, 119,
122-123, 144, 252,446-448; see
also Hebrew; Yiddish
Johnson, Christina 407
Johnson, Ken 86
jokes 188-192
Kaczynski, Mieczysiaw 271
Kageyama, Taro 78
Kalish, Donald 391
Karcevski, Serge 285, 293, 302,
304-305,310
Kasper, Gabriele 60
Katriel, Tamar 67,69,91, 185-187
Keen, Ian 103
Kemme, Hans-Martin 342
Keneally, Thomas 181
Kerry, Tristram 442-443
Kinchin-Smith, F. 316
King, Jonathan 183
Subject and name index 491
kinship 158-160
Kipling, Rudyard 395, 415
Kirby, Michael 419
Kochman, Thomas 3, 68-70, 73, 78-
79,82-85,119,124-125,277
Kolankiewicz, George 192
Konig, Erich 367
Korean 48,105,107,112,396,454
Kucera, Henry 339
Kurokawa, Shozo 13
Labov, William 183-184
Lakoff, George 18, 30, 32, 59, 62,
213
laments 338
Lanham, Betty 80, 86, 126
Latin 7,255-256,270,458,460
Lawson, Henry 167
Lebra, Takie 57, 70, 72-73, 75, 85,
87,91,103,127,153-155,176
Lee, David 350-353
Leech, Geoffrey 15,17,19,21,68,
132, 198-200, 209-210, 277-279,
282
Leibniz, Gottfried Wilhelm 10, 343
Lepschy, Anna and Giulio 257, 265
Levenston, Edward 92
Levinson 30, 67, 108-109,207-209,
254,279,391,397,399-401,408,
432,444-445,454
Levy, Robert 339
Lewis, Clive Staples 414
Lewis, Paul 192
Lithuanian 43, 45
Locke, John 196,341-342
logic 60, 62-63, 207, 391, 397,399-
400,454
Longman (English dictionary) 6,
286-287,289,295,303,317,330
Luo 423, 426-428, 436
Lutz, Catherine 9,54-55, 103, 121,
339
492 Subject and name index
Malamud, Bernard 3-4, 446-448
Manifold, John 182
Manzoni, Alessandro 256-257, 259,
261-262, 266,438-439
Matisoff, James 4, 97, 122-123,249-
251,338,447
Matsumoto, Yoshiko 30, 67, 77-78
McGregor, Craig 170
Mediterranean cultures 47,50,65,
279
Mel'cuk, Igor 270,280
Melluish, T.W. 316
mental verbs 238-240
metalanguage 6-8, 71-72, 130-131,
134, 137, 147-149, 151, 156, 161-
165, 184, 199-202, 212, 218-219,
253, 283,288,291, 314, 337-338,
343-344,403-404,451,453-455
Mezzrow, Milton 183-184
Mickiewicz, Adam 50, 297-298, 312
Miller, Roy 145-146
Minami, Hiroshi 154
Mitchell-Kernan, Claudia 277
Mittwoch, Anita 208-209
Mizutani, Osamu and Nobuko 68,
69,74-75,81,86,93-95, 126,
145-147
modal verbs 235-238
Moiseev, Aleksandr 367
Monahan, Barbara 47, 53
Montague, Richard 391
Montgomery, Lucy 396
Moravcsik, Edith 270
Morgan, Jerry 16,451
Morimoto, Junko 153-155
Morris, Charles W. 15, 16
Morris, Desmond 446
Myerhoff, Barbara 92
Nabokov, Vladimir 285
Nakane, Chie 93, 112, 153
names, see address forms
natural kinds 17-18
natural semantic metalanguage, see
metalanguage
Nesbitt, Edith 408
Nevile, Ann 152-153
Newmeyer, Frederick 18
Niehoff, Arthur 91
Nikolaeva, Tat'jana 342
Nuchelmans, Gabriel 240, 458
O'Grady, John 177-178
O'Neail, Emmie 228
O'Neill, Eugene 459
objectivity 49-50
obligation 157-160,175-176,419-
423
Ochs (Keenan), Elinor 69
OED (Oxford English Dictionary)
177, 180, 360
offers 27-30, 211-212, 215-216, 228
k u d ~ a v a Bulat 393, 435
Olshtain, Elite 77
opinions 31, 41-44, 219
opposite, contrastive 139
orders 30, 39, 151-152, 158-160,
199-202, 205, 215-216; see also
commands; imperatives
Oring, Elliott 185-186
other minds 162-165
Oztek, Piale 97
Paduceva, Elena 17, 252, 363
Palakornkul, Angkab 12
particles 23, 240-243, 341-389;
approximative 354-366, 384-389;
quantitative 345-366, 379-389;
temporal 367-378
Pascal, Fania 405
Pasicki, Adam 367, 371
past 434-439
paternalism 154
Paterson, Katherine 459
Paul of Venice 240
performatives 37, 60, 163, 197-198,
201-202, 208-209, 236, 253-254
Pitjantjatjara 293
Polish 26,71, 103,279,454; advice
31-32; address forms 47-49,51,
107, 112; culture 47-60, 105, 114-
115, 121-122; diminutives 51-53,
122; exclamations 45-46; impera-
tives 30-31, 77, 88, 122; interjec-
tions 23, 286, 290, 292-308,
310-315,318-334,336-337,315;
offers 27-29; opinions 41-44;
particles 23, 344, 362, 367, 370-
389; requests 32-37; speech acts
20-21,27-60, 165; speech genres
188-196; tags 37-41; tautologies
394-395, 402-403, 413, 422-423,
425-426,432,434,436-437,445
politeness 13, 27, 30, 34-36, 40, 43-
44, 52, 56-64, 86, 90, 100, 205,
213; see also address forms;
honorifics
polysemy 11, 350
Pomerantz, Anita 21, 131, 133, 136-
144, 148
practical applications, see intercultu-
ral understanding
pragmaticism 17-18
pragmatics, non-linguistic 6, 131-
136
pragmatics, vs. semantics 1-24, 366,
449-455
praise 68-69, 136, 141-142, 146-147,
232-235
Pride, Janet 283
primitives, semantic see metalan-
guage; universals
Pritchard, Katherine 167
promises 151-152, 198
pronouns 12-14
Psathas, George 254
Subject and name index 493
questions 5-6, 27-30, 32-35, 45-46,
52-53,61-63,75-8,89, 102, 132,
150-151, 199-201,203-204,209-
217,220,224-232,240-241,451;
see also tags
Quirk, Randolph 274
Raphael, Frederic 459
Rath, Rainer 393
Rayner, Clare 286
realism 404-405
reciprocity, see obligation
reduplication 22, 255-284
Reisman, Karl 85
relevance 220
Renwick, George 3, 136, 169, 173,
183
repetition, clausal 258-263, 268-270
reprimands 223-224
requests 25, 31-37, 39, 52, 60-62,
76-78, 138, 151-152, 159, 193-
195,201-203,215-217, 228
respect, see politeness; formality
restraint 45, 53, 74-76, 79, 83, 85-
87, 99, 128
return, of compliment 142-143
Rintell, Ellen 26, 152
Rosaldo, Michelle 152, 339
Ross, John R. 197, 238
Rosten, Leo 285, 294,308-310,317,
371, 373
Russell, Bertrand 14
Russian 7, 17, 30, 71, 279-280, 454;
address forms 57-59; culture 2,
105, 110, 114, 129-130; impera-
tives 77, 88; interjections 23, 285,
293,296,300-302,310-315,322-
323, 326-330, 334-336, 338-339;
particles 342, 367, 393; requests
204; tautologies 392-394, 425-
426,435,445
Rymkiewicz, Jaroslaw 413
494 Subject and name index
Sabra (Israeli) 185-186
Sacks, Harvey 254
Sadock, Jerrold 23,200-201,214,
229, 356, 363-364, 366
Sakurai, Kaoru 157
Sansom, Basil 63, 103
Sapir, Edward 279,282, 365
sarcasm 38
Schachter, Josef 197
Schegloff, Emmanuel 254
Schenkein, Jim 148, 254
Schiffrin, Deborah 68,69,92,119
Schreiber, Paul 207
Schrett, J6zef 59
Schuchardt, Hugo 19
Schunk, Dale 25
Searle, John 59-62,67, 151,201-
202, 214, 454
Sekiguchi, Isugio 393
self-assertion 72-88
self-restraint, see restraint
semantic metalanguage, see
metalanguage
semanticism 18
Serbo-Croatian 30, 435
sex differences 58, 136, 145-146,
166, 168, 172, 175, 183
Shakespeare, William 118,313
shame 306-307
Sherzer, Joel 150
Shetter, William 367
Shin, Gi-Hyun 396
Shopen, Timothy 217, 220
silence, calls for 293-296
Simpson, Jane 64, 269, 279
sincerity 70-71, 104, 115-121, 133,
135, 186; see also directness
SJP (Polish dictionary) 193, 294,
297-300, 306, 312, 318-321, 323-
327, 331-333, 376, 380-382, 385,
389
Skinner, Quentin 165
Skorupka, Stanislaw 193
Slavic languages 47, 50-51, 56,
107,114,118,279
Smith, Gerda 375
Smith, Hedrick 53, 86
Smith, Robert 57,74,91, 153
sociology 254
Sohn, Ho-min 48, 69-70, 105-
107, 111
Solomon, Robert 339
sound symbolism 294,315-317,
337-338
Spanish 30, 51, 402-403, 431-434
speech acts 17, 21-22, 25-65,
149-196, 197-254
Sperber, Dan 449
spitting 310-313
Spitzer, Leo 282
spontaneity 57, 70-71, 80-83, 86
Sprachgeist 282
SRJa (Russian dictionary) 300,
327, 329, 335
SSRLJa (Russian dictionary) 296,
300,305,311-312,322,329,
335
Stein, Gertrude 442
Stevenson, Burton 104,401,411
Stoics 197, 240
Sudnow, David 254
suggestions 27, 94, 198, 211-212,
214-216, 219
superlatives 270-276
Suslov, 11'ja 304
Suzuki, Takao 12,72-73,85
swearing 3, 35, 169, 187
Swedish 307,313-314
syntax 197-198, 209-210, 245-
246, 253, 403-404
Szelburg-Zarembina, Ewa 307,
320-321
tags 31, 37-42, 224-232
Tamil 444-445
Tamori, Ikuhiro 78
Tamura, Kyoko 120-121, 127-128
Tannen, Deborah 1,67,69,79,
97-99
tautologies 23, 391-455; nominal
391-431; semantic invariants
439-444, verbal 431-444
taxonomy, folk 150
Taylor, Brian 169
Thai 12, 48, 88
thanks 126, 158
threatening 197-198
tolerance 398, 405-410, 421
Tolstoy, Alexei 296
Tolstoy, Leo 329
Tolstoy, Nikolai 443
Traugott, Elizabeth 367
Triandaphyllidis, Manolis 258
Triandis, Harry 47,96,99, 108
Trilling, Lionel 115-119, 120
truth 103-104, 124, 186, 397, 448
Turkish 12
tum-taking 80-82
Veda, Keiko 145
understatement 44-45, 135, 276-280
ungrammaticality, see grammaticality
universals 9-15, 59-62, 67-69, 71-72,
270,286.289,302,392-397,403,
432, 444, 446, 448, 453-455
Vniversite de Paris-7 342
upgrade 138-139
urging 298-299
Vrmson, James O. 238
values, cultural 21,37,47-65,67-130
Vassiliou, Vasso 96, 99
Verschueren, Jef 152
Volksgeist 282
Volosinov (pseudonym of Baxtin)
457
Vossler, Karl 282
Subject and name index 495
Wajda, Andrzej 36
Waley, Arthur 427
Walmatjari 21, 158-160,454
Walters, Joel 26, 152
Wannan, Bill 182
Ward, Russel 177, 182
warmth, see cordiality
warning 152, 197-198
Waterhouse, John 367
Weber, Max 104
Webster's (English dictionary) 317,
360
Weydt, Harald 342
whimperatives 2, 77-78, 200-207,
209-218
Whorf, Benjamin Lee 282
Wierzbicka, Anna 3, 7-8, 15, 17, 32,
44,49,51,55,57,67,69,86,91,
105-107, 121-122, 128, 133, 136,
153, 155, 157, 160, 162, 165,
172,181,193,195,200-202,217,
221, 243, 268, 272, 277, 287-289,
338-339, 342-344, 432, 459
Wilkes, Gerald 166-167, 178, 183
Wilkins, David 270, 344
Williams, Tennessee 241, 459
Williamson, David 35, 39-40, 50,
178,205,460
Wilson, Deirdre 450
Winter, Werner 460
Wittgenstein, Ludwig 15, 150, 197
Wolfe, Bernard 183-184
Yiddish 4,21,23,71, 122,249-252,
285-286,294,308-310,315-318,
338,370,371,447,454
Yolngu 158
Zapolska, Gabriela 36-37
Zasorina, L.N. 339
Zipf, George 11
Index of words and phrases
This index lists words and phrases explicated or discussed in the book; explica-
tions are indicated by bold page numbers. Languages are abbreviated as
follows: Da(nish), Ch(inese), Fr(ench), Ge(rman), Gr(eek), He(brew), It(alian),
Jp(Japanese), Jv(Javanese), Ko(rean), La(tin), Pi(tjantjatjara), Po(lish), Ru(ssian),
Se(rbo-Croat), Sp(anish), Sw(edish), Wa(lmatjari), Yi(ddish). All items are listed
in the order of the English alphabet.
a joke is a joke 410, 422
a kiss is just a kiss 415, 418, 441
a man is a man 391, 411,412-413,
424,439
a party is a party 403, 408-409, 417,
418-419, 441
a picnic is a picnic 408, 409
a point is a point 415, 418, 441
a promise is a promise 23, 391,419-
420,421,441
a rose is a rose is a rose 442
a steak is a steak 403,416, 419
aah 325
about 23, 344, 355-357,358,361-
364, 366
ach (aah, Po) 323-326,324
adagio adagio (very slowly, It) 255,
264-265
aga (aha, Ru) 327-328,329-330
agreement 137; scaled-down 140
ah 286-287, 289, 317-318, 325, 339
aha 285,330
aha (aha, Po) 326-332,327, 337
ahoj (ahoy, Ru) 293
aizuchi (turn-taking, Jp) 81
ale co S to S (what has been has
been, Po) 436-437
almost 23, 344, 361-365,366,387,
388
already 1,23, 344,367-368, 369-373
amae (love, Jp) 155
announce 164
apologise 156, 453
application 194, 195-6
appreciation 145
approximately 23, 359, 358-360,
362, 364-365
around 23, 344, 355-364, 358, 366
as many as 345
as much as 345,381
ask 5-6, 32, 60, 159, 160-161, 195,
202, 213
at least 23, 354
at most 354
atashi (I, Jp) 13
au (yoo-hoo, Ru) 300
ax (aah, Ru) 326, 339
az (as much as, Po) 380,381
bark 163
bastard 1, 3, 119, 169, 187
~ (fall, Po) 290
because 208
believe 42-43, 238-239
bella bella (very beautiful, It) 22,
256-258, 265, 278
blisko (nearly, Po) 386-387, 388-389
bloody 3, 55, 169
blue 1
boast 163-164
498 Index of words and phrases
boku (I, Jp) 13
boys will be boys 391-394, 398-403,
407,408,410,424,440-441,444-
445, 449-450
bugger 2-3, 169
bullshit 119, 169
business is business 391, 398,404,
450
bygones, see let
caffe caffe (real coffee, It) 265, 267-
268
can you 25-26,61,89,204
can't you 35, 38,207, 227-228,229
cat 17
chat 171, 172
chiack 165-170,168,173,175,177,
187
cii (shh, Po) 293-294, 295
cip-cip-cip (here chickens, Po) 292
closeness 108-111, 109
co to (what will be
will be, Po) 402-403,432
co by/o to by/o (what has been has
been, Po) 434-436
co X to X (an X is an X, Po) 395,
413
come in, come in! 258, 260, 262,
263, 269
command 34, 150-151, 199
complain 164, 181,453
compromise 48-49
conclude 241-242
confess 161-162
cooee 301
cordiality 122
could you 38, 53,88-89,96,201,
204, 213
country 49
criticise 213
cto bylo to bylo (what has been has
been, Ru) 435
cup 17
curses, Jewish 123
dakara (of course, Jp) 429
damn you 249,250-251
darling 1
dear sir 118
demo (something, Jp) 94-95
did you know 221-222
ding shi ding (ding is ding, Ch) 426-
427
disagreement 137
do it 96,205
do you know 220-222, 221
dob 177-180,179
dobrze? (okay?, Po) 38
doch (indeed, Ge) 393
doggie 1, 50, 55
don't tell me 222, 223
dopiero (only, Po) 376-378,377,
380
Uoke,Po) 188
downgrade 137, 141-142
dozens 165, 183-185, 184
du (you, Ge) 13
dugri (straight, He) 165, 185-188,
187
east is east 395, 404, 413, 414, 426-
427
either he will or he won't 400-402,
433,447
emotional 53-55, 121
empathy 127
enough is enough 410, 439, 442, 450
enryo (self-restraint, Jp) 74-78, 76,
95, 99
-est 272-276
est' (be, Ru) 393-394
e{ok-e{ok (dissimulation, Jv) 100-
101, 102
exactly 359
excuse me 89, 108
fe (yuk, Po) 23, 306-307,308,314
fear (I fear) 239
feh (yuk, Yi) 308-309, 310, 314, 316
feu (phew, Gr) 315-316
for God's sake! 143-144
forbid 152
frankly 207-208,210
friendliness 87
fu (yuk, Po) 23, 286, 302-306, 303,
313-314,316
fu (yuk, Ru) 302, 304-306,305,313-
314, 338-339
fy (yuk, Sw/Da) 307-308, 313-314,
338
gather 241-242, 238, 239-240
gee 244, 286-287, 290, 330
go jump in the lake 246, 247
good morning 246
good night 246
good-bye 1, 246
grandstanding 68, 79, 84
guess 43
hallo 293
hardly 382
harmony 114-115
have a nice day 86-87
hej (hey, Po) 293, 298, 299
hejta (gee, Po) 292
hejie (come on, Po) 293, 299
hell! 290-291
hito wa hito da (a person is a
person, Jp) 430-431
hop hop (ahoy, Po) 293,300, 301
hop, hopla (jump, Po) 290, 293,298
how about 27,62,215-218,216,245
how are you 116, 118, 132, 133-136;
replies 134-135
how dare you 246, 247
how do you do 62
Index of words and phrases 499
how many times have I told you?
223, 224
how nice 45-46, 287
huzza (sic-em, Po) 292
hyc (jump, Po) 290
I 12-14
ie (house, Jp) 1
-ie 55-56
inflexible 49
inform 177-179
informality 111-113, 112
intimacy 5 ~ 106-108
invite 211-212, 215
is it 226, 227
isn't it 45-46, 225, 226
-issimo/a (-est, It) 270-271, 272,
276, 278-279
-ja (certainly, Ko) 396
japirlyung (request, Wa) 159, 160
jeszcze (still, Po) 371-375, 372
jinjinyung (order, Wa) 160
joke 188-191,192
just 23, 344, 349-353, 354
jui: (already, Po) 371-376, 373
kansha suru (thank, Ja) 157
kawaJ (joke, Po) 165, 188-192, 191
kici-kici-kici (here kitty, Po) 292
kids are kids 391, 405-406, 407,
411,440-441,450
kimi (you, Jp) 13, 48
kompromis (compromise, Po) 48-49
kraj (country, Po) 49
kudasai (please, Jp) 78
kure (please, Jp) 78
least, at 354
ledwie (just, Po) 382-385, 384
let bygones be bygones 434, 441
letter 194
list (letter, Po) 194
500 Index of words and phrases
many, as 345
materit'sja (mother-swear, Ru) 17
merely 23,348-351,349
molto (very, It) 257, 264-266, 268,
278
more 372
more or less 345
most (Adv.) 22,272-275,276,278-
279
most, at 354
mostly 344
mozzies 55, 56
much, as 345
mug 17
must 236-238, 237
na (take it, Po) 293
nanren shi nanren (men are men,
Ch) 423, 424
naplevat' (spit, Ru) 310-312
narod (nation, Po) 49-50
nearly 365, 366
nedaru (ask for, Jp) 155
neri neri Get black, It) 256, 267,
269, 278
nezumi (rat/mouse, Jp) 17
niemal (almost, Po) 385-387, 386
nieomal (almost, Po) 385
(inflexible, Po) 49
niezlomny (unbreakable, Po) 49
no 92-93
no less 355
no more 355
noch (yet, Ge) 375-376
nu (interj, Yi) 285-286
nuie (come on, Po) 293, 298, 299
o (oh, Po) 326, 333, 334
o (oh, Ru) 339
o malo nie (only just, Po) 362, 384,
385, 387
och (ooh, Po) 323-326,324, 337
offer 211-212, 215
ogo (oho, Ru) 334, 335-336
oh 325,339
oh my God 223, 243, 244-245
oh-oh 332, 334
oho (oho, Po) 326-327, 331-334,
332, 337
oj (oy, Po) 318-322,323,325,327
oj (oh dear, Ru) 322, 323, 339
ojej (oh no, Po) 320,321
OK 38, 97, 231,232
omal (almost, Po) 385
omoiyari (empathy, Jp) 87
only 23,344,346,347-351,355,
370, 378-381
ooh 325
oops 287, 289, 290
opposite, contrastive 139
order 17, 34, 151-154, 158-159, 160,
164-165, 198-201, 202
ore (I, Jp) 13
otaku (esteemed house, Jp) 1
ow/ouch 286-289
ox (ooh, Ru) 326, 339
oy 285,317-318
oy vay (oh no, Yi) 338
pan/pani (Mr/Mrs, Po) 48, 56-59,
107
pfui 286,313-315,338
phew 286-287, 304, 313-315, 339
pismo (official letter, Po) 194
please 201, 203-204
pledge 17
podanie (application, Po) 165, 192-
196, 195
pooh 286,315,339
praise 453
prawda (true, Po) 38-40, 42
prawie (almost, Po) 362, 385-388,
386-387
prawo prawem (the law is the law,
Po) 422-423,
precz (go away, Po) 293, 296, 297
privacy 47
problem6n/-azo (big problem, Sp) 1
promise 17, 149, 151, 164,453; see
also a promise is a promise
p r o s z ~ (please, Po) 28, 34, 60, 195
prr (whoa, Po) 292
psst 295-296, 337
pst (psst, Po) 293-296, 295
pst (pah, Ru) 296
que sera sera (what will be will be,
Sp) 402, 431-432, 433-434
question 5-6, 149-151, 199
r z ~ ei (I advise you, Po) 31-32, 60
rapping 3, 68, 79, 150
rather 43-44, 276-277
reckon 42-43
referent shift 137, 142
reprimand 154, 453
request 17,32,149,151,161,194-
195, 198-202
return 143
reveal 161-162
rieeo rieeo (very rich, It) 267
roughly 23, 360, 361
rubbish 119, 187
sake (rice whisky, Jp) 17
same here 50
satosu (warn, Jp) 153, 155-156
say 11
scab 170
schon (already, Ge) 375-376
sensei (teacher, Jp) 48
shout 170, 173-174, 175, 176-177
showboating 68, 79, 84
sincerity 71, 115-121, 117, 121
sio (shoo, Po) 293
sit down! 27
so 240-242, 241
sook 182
sort of 44
Index of words and phrases 501
sounding 150, 183
spin a yarn 171
still 23, 344, 367-368, 369-372
sto je bi/o bi/o je (what has been has
been, Se) 435
stylin' out 79, 84
subito subito (at once, It) 257-258,
266,267,271
suggest 202, 211-212, 215
sumimasen (indebted, Jp) 157
super/ativo (superlative, It) 270-271
sza (hush, Po) 293, 294-295
tanomu (ask, Jp) 155
tas-tas-tas (here rabbits, Po) 292
tell 34, 108-110
tell on 179
tell you what 219, 220
t'fu (spit, Ru) 23, 310-313, 311, 314,
339
tfu (spit, Po) 23, 310-313, 312, 314
thank 156, 157, 262, 453
thanks 1
that's that 425, 439-441
the law is the law 404, 419-420,421
then 240-242
think 41-43
threaten 153, 156, 163, 198, 201,
453
thunder 163-164
tju-tjuu (here doggie, Pi) 293
-to (too, Ko) 396
tpru (whoa, Ru) 293
truth 103-104, 125
tu (you, Fr) 13, 48
ty (you, Po) 48, 56-59
ty (you, Ru) 58
tylko (only, Po) 379, 380-381
uczuciowy (emotional, Po) 54
upgrade 137, 138
understand 238
ux (oof, Ru) 335
502 Index of words and phrases
ver (become, Yi) 250
very 256-257, 263-269
vow 17
vy (you, Ru) 58
war is war 23, 394, 400-401, 403-
404,405,420,422,440-441,444-
445
wara (keep away, Po) 293, 296-299,
298
warmth 87
warn 149, 152-154, 153, 156, 163-
165, 198,201,453
watakushi (I, Jp) 13
watashi (I, Jp) 13
what's done is done 437
whatever will be 440-441, 431-433
whinge 180, 181-182
who's talking about 224
why can't you 38, 229, 230-231
why do X 213, 214-215
why don't you 32-35, 38, 60, 62,
210-213, 212, 215-216, 218, 245,
449, 452
why not 214, 215
w z ~ (I see, Po) 39
wii (whoa, Pi) 293
will 235-237, 236
will you 27, 32, 35, 38-39, 60, 204,
206, 207, 227-229, 228
wio (gee-up, Po) 292
wista (haa, Po) 292
won (get out, Po) 293, 296,297
won't you 27,38-39,227-228,229
worries, no 56
would you 2,27,29,38,51,53,62,
88-89,96, 203-204, 206, 207,
449, 452
wow 286-287, 289, 290, 330,459
wy (you, Po) 58-59
yarn 170-171, 172-173, 175, 177
yellow 1
yet 23, 344, 367-369,368,371
you 11, 12-14,47-48,56-59
you beauty! 234,235
you X! 233-234, 235
yuk 1,23,288,302-303,304,313-
315,339,459-460
zaledwie (only, Po) 381-383,382
zart Uoke, Po) 188-189
zarty zartami Uokes are jokes, Po)
422
ze (indeed, Ru) 393
zol (let/should, Yi) 251